Beam monitoring method for particle beam therapy system, beam monitoring program, information processing system, and particle beam system

Abstract

To adjust a beam orbit or a beam position of a charged particle beam in an accelerator.SOLUTION: The present disclosure relates to a beam monitoring method 100 for a particle beam therapy system 160. The particle beam therapy system 160 comprises an accelerator 170. The beam monitoring method 100 is executed by an information processing system 101. The beam monitoring method 100 includes a beam position information acquisition step 102 and a parameter acquisition step 103. The beam position information acquisition step 102 is the step of acquiring information of beam positions in beam monitors 181-184. The parameter acquisition step 103 is the step of acquiring parameters, which may be used for adjusting a beam orbit of a circulating charged particle beam 171 based on the acquired information of the beam positions, to be set to electromagnets 191-194.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to a technique for monitoring a charged particle beam in a particle beam therapy system that irradiates a diseased area with a charged particle beam for treatment. [Background technology]

[0002] A particle beam therapy system typically includes at least a charged particle beam generator, an accelerator, and an irradiation device. The charged particle beam generator injects the generated charged particle beam into the accelerator. The accelerator (e.g., a synchrotron) imparts a desired amount of energy and properties to the charged particle beam injected from the charged particle beam generator. The accelerator emits the charged particle beam that has been given the desired amount of energy and properties. The irradiation device irradiates the charged particle beam emitted from the accelerator to an affected area to perform treatment. An example of a prior art for such a particle beam therapy system is disclosed in Patent Document 1. In the prior art disclosed in Patent Document 1, a position monitor or a beam position monitor measures the position of a charged particle beam extracted from an accelerator. Also, in the prior art disclosed in Patent Document 1, various electromagnets (e.g., steering electromagnets) adjust the position of the charged particle beam extracted from the accelerator. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 5707524 Summary of the Invention [Problem to be solved by the invention]

[0004] A particle beam therapy system is intended to irradiate a diseased area with a charged particle beam having a desired energy amount and properties. Here, if the charged particle beam deviates from a desired beam trajectory (design beam trajectory) or a desired beam position in the accelerator, the energy amount and properties of the charged particle beam also deviate from those desired. In fact, due to various causes, the charged particle beam may deviate from a desired beam trajectory (design beam trajectory) or a desired beam position in the accelerator. If the energy or properties of the charged particle beam deviate significantly from what is desired and fall outside the normal range, operators may be forced to shut down the particle therapy system or accelerator, and may be forced to change the treatment or operational schedule by adjusting the accelerator. In addition, if the charged particle beam deviates from the desired beam trajectory (design beam trajectory) or the desired beam position in the accelerator, the efficiency of the charged particle beam being extracted from the accelerator may also decrease. If the efficiency of the charged particle beam being extracted from the accelerator decreases, the treatment time using the particle beam therapy system may be extended. Here, if the charged particle beam deviates from the desired beam orbit (design beam orbit) or desired beam position within the accelerator, there is a risk that the energy amount and properties of the charged particle beam will deviate from those desired. However, with the prior art disclosed in Patent Document 1, in which the position of the charged particle beam extracted from the accelerator is adjusted by controlling the excitation current amount of various electromagnets (e.g., steering electromagnets), it is not possible to bring the energy amount and properties of the charged particle beam closer to the desired state. Furthermore, in the prior art disclosed in Patent Document 1, the position of the charged particle beam is adjusted after it is extracted from the accelerator, and therefore it is not possible to improve the efficiency of extracting the charged particle beam from the accelerator. In order to prevent the charged particle beam in the accelerator from deviating from the desired beam trajectory (design beam trajectory) or the desired beam position, the accelerator operator may adjust parameters related to the control of the accelerator (e.g., parameters of the electromagnets in the accelerator). However, manually adjusting all of the parameters related to accelerator control places a burden on the accelerator operator.

[0005] In view of the above, it is an object of the present disclosure to adjust the beam orbit or beam position of a charged particle beam in an accelerator. It is also an object of the present disclosure to assist an accelerator operator, etc., when adjusting the beam orbit or beam position of a charged particle beam in an accelerator. [Means for solving the problem]

[0006] In order to achieve at least one of the above objects, the present disclosure may have the following features, for example. One aspect of the present disclosure is a method for monitoring a beam in a particle therapy system. The particle beam therapy system includes an accelerator for accelerating a charged particle beam. The accelerator includes a circulation path for circulating the charged particle beam, a plurality of beam monitors arranged in sequence along the circulation path, and a plurality of electromagnets arranged in sequence along the circulation path. The beam monitor acquires beam position information, which is the position of the charged particle beam when the circulating charged particle beam passes near the beam monitor. The electromagnet adjusts the beam trajectory of the charged particle beam when the circulating charged particle beam passes near the electromagnet. The beam monitoring method is executed by an information processing system. The beam monitoring method includes a beam position information acquisition step and a parameter acquisition step. The beam position information acquisition step is a step of acquiring information on the beam position in each of the beam monitors. The parameter acquisition step is a step of acquiring parameters to be set in each of the electromagnets, which can be used to adjust the beam trajectory of the circulating charged particle beam, based on the information on the beam position acquired by the beam position information acquisition step. Effect of the Invention

[0007] As described above, the present disclosure can adjust the beam trajectory or beam position of a charged particle beam in an accelerator. Also, as described above, the present disclosure can assist an accelerator operator or the like in adjusting the beam trajectory or beam position of a charged particle beam. Therefore, the present disclosure can provide a desired energy amount and properties of a charged particle beam. The present disclosure can also improve the availability and efficient operation of a particle beam therapy system. Furthermore, the present disclosure can reduce the burden on operators, etc.

[0008] A program or information processing system that realizes the same processing as that realized by the above beam monitoring method can also obtain the same effect as the above beam monitoring method. Furthermore, if it is in the form of a program, the cost can be reduced in many cases. With a program, design changes regarding the processing can be easily made. Other features that the present disclosure may have and the effects corresponding to these features will be disclosed in this specification, the claims, or the drawings. [Brief description of the drawings]

[0009] [Figure 1] 1 illustrates a functional configuration of an embodiment of the present disclosure. [Diagram 2] 1 illustrates a system configuration according to an embodiment of the present disclosure. [Diagram 3] The beam position (BP) and deviation are shown. [Figure 4] Indicates one record in a database (DB). [Diagram 5] The probable cause vector is shown. [Figure 6] 1 illustrates a computer architecture for implementing an embodiment of the present disclosure. [Figure 7] The process of cause estimation, parameter adjustment (creation), and warm-up operation judgment is shown. [Figure 8] 4 shows a rule-based cause inference process. [Figure 9] This shows the process of inferring the cause by searching the database. [Figure 10] The input and output of the cause estimation model are shown. [Figure 11] This shows machine learning of a cause inference model. [Figure 12] 4 shows a cause inference process using a cause inference model. [Figure 13] This shows the parameter adjustment (creation) process using COD correction. [Figure 14] This shows the parameter adjustment (creation) process by DB search. [Figure 15] The input and output of the parameter acquisition model are shown. [Figure 16] 1 shows machine learning of a parameter acquisition model. [Figure 17] 3 shows a parameter adjustment (creation) process using a parameter acquisition model. [Figure 18] The process of creating a parameter set (three-dimensional array) is shown below. [Figure 19] 4 shows a display screen that displays parameters and presumed causes. [Figure 20] 4 shows a warm-up operation determination process based on cause estimation. [Figure 21] This shows the warm-up operation determination process by DB search. [Figure 22] The input and output of the warm-up operation judgment model are shown. [Figure 23] This shows the machine learning of the warm-up operation judgment model. [Figure 24] 4 shows a warm-up operation determination process using a warm-up operation determination model. [Diagram 25] 4 shows the adjustment time determination process. [Figure 26] The relationship between the estimated BP time series information and the parameter adjustment timing is shown. [Figure 27] 1 shows the process of generating a regression analysis model for estimating BP time series information. [Figure 28] The estimation process of BP time series information using a regression analysis model is shown. [Figure 29] The input and output of the BP time series estimation model are shown. [Diagram 30] We present the machine learning of the BP time series estimation model. [Diagram 31] 1 shows a process for estimating BP time series information using a BP time series estimation model. [Diagram 32] 1 illustrates remote monitoring. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, the embodiments of the present disclosure will be described in detail with reference to the drawings. Note that the embodiments described below do not limit the disclosure according to the claims, and all of the elements and combinations thereof described in the embodiments are not necessarily essential to the solution of the present disclosure. The following description and drawings are examples for explaining the present disclosure, and are omitted and simplified as appropriate for clarity of explanation. The present disclosure can be implemented in various other forms. Unless otherwise limited, each component may be singular or plural. The position, size, shape, range, etc. of each component shown in the drawings may not represent the actual position, size, shape, range, etc. in order to facilitate understanding of the invention. For this reason, the present disclosure is not necessarily limited to the position, size, shape, range, etc. disclosed in the drawings. Each of the systems, devices, or units of the present disclosure may be integrated as a single piece of hardware, or may be divided into multiple parts that work together to perform their roles. Several systems, devices, or units may be integrated as hardware. Each of the systems, devices, or units may be realized by having a computer execute software (programs) (as in FIG. 6). Some of the functions of the system, device, or unit may be realized by hardware (e.g., hardwired logic or FPGA (Field Programmable Gate Array)), and the remaining functions may be realized by executing software (programs). All of the functions of each of the systems, devices, or units may be realized by hardware. Some or all of the steps of the processes shown in the flowcharts and the like described in this disclosure may be realized by hardware. The program of the present disclosure may be included in the concept that includes software in general, such as software and hardware resources working together to construct a specific information processing device or its operating method according to the purpose of use. In other words, the program of the present disclosure is not limited to a specific type or form of program. In addition, the program may be initially recorded in a compressed format. The same reference numbers are used in multiple drawings to indicate similar things. In the drawings showing the flowcharts described below, rectangular boxes indicate processing steps, and diamond-shaped boxes indicate conditional branch steps. In the drawings showing the flowcharts, "step" is abbreviated as "S." Also, the screen aspects shown in the drawings described below are examples, and are not limited to these.

[0011] 1. Functional configuration of the embodiment of the present disclosure Fig. 1 shows a functional configuration of an embodiment of the present disclosure. Note that not all of the functional configurations shown in Fig. 1 are essential. 1 shows a beam monitoring method 100 according to the present disclosure, an information processing system 101 that executes the beam monitoring method 100, and a particle beam therapy system 160 that is a target of monitoring by the beam monitoring method 100. Here, the particle beam therapy system 160 and the information processing system 101 are collectively referred to as a particle beam system 105. The particle beam therapy system 160 includes an accelerator 170 for accelerating a charged particle beam 171 (hereinafter, the charged particle beam may be referred to as a "beam"). Although not shown in FIG. 1, the particle beam therapy system 160 also includes a charged particle beam generator that injects the beam 171 into the accelerator 170, and an irradiation device (irradiation system) that irradiates the affected area with the beam 171 extracted from the accelerator 170. The type of charged particles constituting the beam 171 is not limited, and may be, for example, hydrogen, helium, or carbon. The particle beam therapy system 160 may also include a beam transport system that transports the beam 171 extracted from the accelerator 170 to the irradiation device (irradiation system). The accelerator 170 includes a circular path 172 for circulating the beam 171. In Fig. 1, the circular path 172 is shown to be composed of four straight path segments and four curved path segments, but the shape of the circular path 172 is not limited to the shape shown in Fig. 1. The accelerator 170 includes a plurality of electromagnets 191-194 that are installed in sequence along the circular path 172, and a plurality of beam monitors or beam position monitors (hereinafter, the beam monitors or beam position monitors may be referred to as "BPMs") 181-184 that are installed in sequence along the circular path 172. In Fig. 1, the accelerator 170 is shown equipped with four electromagnets and four BPMs, but the number of electromagnets and the number of BPMs in the accelerator 170 are not limited to four. The BPMs 181 to 184 acquire information on the beam position (hereinafter, the beam position may be referred to as "BP") of the charged particle beam 171 when the beam 171 orbiting the circulation path 172 passes near the BPMs 181 to 184. The BP in the BPM may be the position of the beam 171 in a coordinate system intersecting the traveling direction of the circulation path 172 when the circulating beam 171 passes near the BPMs 181 to 184. In other words, the BP can be treated as the coordinates of a point in a one-dimensional coordinate system (in the distance direction from the center of the plane drawn by the circulation path 172) or the coordinates of a point in a two-dimensional coordinate system (consisting of the distance direction and the perpendicular direction (up and down direction)). For example, the BPMs 181 to 184 include several electrodes close to the circulation path 172. When the beam 171 passes near these electrodes, a potential is applied to the electrodes. Since the potential of each electrode is determined according to BP, the BP can be determined by measuring the potential of each electrode. In this manner, the BPMs 181 to 184 are non-destructive monitors and do not damage the circulating beam 171. In the accelerator 170, the BP acquired by the BPM can be handled as one-dimensional or two-dimensional point coordinates, making it possible to simplify data handling. In addition to the distance component and vertical (up-down) component, BP may also include a component in the traveling direction of the circular path 172. If information on such components is used together with time information, it is expected that auxiliary information on the speed of the beam 171 can be obtained. The electromagnets 191-194 adjust the beam orbit 173 when the circulating beam 171 passes near the electromagnets 191-194. The electromagnets 191-194 apply a magnetic field having an excitation amount B to the circulating path 172 according to the magnitude of the current applied (current amount I). The larger the excitation amount B, the greater the force generated in the circulating beam 171. In FIG. 1, electromagnets 191 to 194 are bending electromagnets. Incidentally, the electromagnet having a steering function for adjusting the beam trajectory 173 or BP is not limited to a bending electromagnet. For example, such an electromagnet may be another type of electromagnet, such as a steering electromagnet applied to a linear path portion. For example, the bending electromagnet may be specialized for the function of deflecting the beam trajectory 173, and a steering electromagnet provided in the linear path portion may be modified to perform the steering function. In this way, the configuration of the electromagnets can be flexibly determined. The beam monitoring method 100 executed by the information processing system 101 includes a BP information acquisition step 102 and a parameter acquisition step 103. The BP information acquisition step 102 acquires information on the BP in each of the BPMs 181-184. In the parameter acquisition step 103, parameters to be set for each of the electromagnets 191-194 that can be used to adjust the beam orbit 173 of the circulating beam 171 are acquired based on the BP information acquired in the BP information acquisition step 102. Here, the parameter acquisition method may be a calculation method, a method of acquiring the parameters by searching a database (hereinafter, the database may be referred to as "DB"), a method of acquiring the parameters from the output of a trained model, or other methods.

[0012] Since the beam monitoring method 100 disclosed herein has the functional configuration described above, it can provide the effects shown in the above-mentioned [Effects of the Invention].

[0013] 2. System configuration of the embodiment of the present disclosure FIG. 2 shows an information processing system 101 (or an information processing system 101 which is one aspect of the present disclosure) that executes a beam monitoring method 100 which is one aspect of the present disclosure, and a system configuration for realizing a particle beam therapy system 160 which is the subject of monitoring by the beam monitoring method 100 (or the information processing system 101).

[0014] The particle beam therapy system 160 includes an accelerator 170, an accelerator control device 240, a data accumulation server 250, a room temperature sensor 281, a cooling water temperature sensor 282, an installation surface position sensor 283, an outside air temperature sensor 284, and a sensor data collection device 280. Among the components included in the particle beam therapy system 160, at least the accelerator 170, the room temperature sensor 281, and the installation surface position sensor 283 are located in a building 270 having a room in which the accelerator 170 is installed. The accelerator control device 240, the data accumulation server 250, and the sensor data collection device 280 may be located in the building 270 or in a location different from the building 270. The accelerator 170 is the same as that described with reference to FIG. 1, so a description of the accelerator 170 will be omitted here. The accelerator controller 240 controls the accelerator 170. To control the accelerator 170, the accelerator controller 240 selects parameters (for example, parameters for the amount of current I of the electromagnets 191 to 194 and parameters related to the mode of warm-up operation or cooling operation) and applies them to the accelerator 170. To select the parameters and apply them to the accelerator 170, the accelerator controller 240 may be provided with an interface that can be handled by an operator or the like. The accelerator controller 240 includes a parameter storage device 241. The parameter storage device 241 stores parameter sets that can be applied to the accelerator 170. The accelerator controller 240 may select appropriate parameters from the parameter sets according to, for example, the amount of energy to be given to the beam 171 and apply them to the accelerator 170. The parameter sets to be applied to the electromagnets 191 to 194 will be explained in more detail later with reference to FIG. 18. The accelerator control device 240 also collects various logs of the accelerator 170. The accelerator control device 240 collects, for example, BP information obtained by the BPMs 181-184, electromagnet information consisting of parameter information of the current amount I set in the electromagnets 191-194 or the excitation amount B in the electromagnets 191-194, and system information logs consisting of the date or time associated with each piece of information, the elapsed time after the start-up or restart of the particle beam therapy system 160, and the elapsed time after the building 270 was constructed. The accelerator control device 240 includes a log temporary storage device 242. The log temporary storage device 242 temporarily stores the logs of the BP information, electromagnet information, and system information collected by the accelerator control device 240. The accelerator control device 240 transmits the BP information, electromagnet information, and system information temporarily stored in the log temporary storage device 242 to the data accumulation server 250. The data accumulation server 250 accumulates the BP information, electromagnet information, and system information transmitted from the accelerator control device 240. The sensor data collection device 280 collects measurement values ​​of various sensors related to the accelerator 170. Among the various sensors, a room temperature sensor 281 measures the room temperature of the room in which the accelerator 170 is stored. A cooling water temperature sensor 282 measures the temperature of the cooling water used in the accelerator 170. An installation surface position sensor 283 measures the inclination and shape of the floor on which the accelerator 170 is installed. An outside air temperature sensor 284 measures the temperature of the outside air surrounding the building 270.

[0015] The information processing system 101 includes a beam position information acquisition unit (BP information acquisition unit) 211, an electromagnet information acquisition unit 212, a sensor data acquisition unit 213, a system information acquisition unit 214, a cause estimation unit 221, a parameter acquisition calculation unit 222, a warm-up operation determination unit 223 (as will be shown later as a modified example, a determination of not only a warm-up operation but also a cooling operation may be performed), an adjustment timing determination unit 224, a display output control unit 231, and a parameter setting unit 232. (In FIG. 2 and FIG. 32 described later, the "unit" part is omitted.) Of these, the BP information acquisition unit 211 can execute the BP information acquisition step 102 in FIG. 1. Also, the parameter acquisition calculation unit 222 can execute the parameter acquisition step 103 in FIG. 1. The BP information acquisition unit 211 acquires BP information from the data accumulation server 250. In addition, the BP information acquisition unit 211 judges whether or not the deviation of the beam trajectory 173 from the desired beam trajectory (design beam trajectory) falls outside the adjustment-free range, based on the acquired BP information. FIG. 3 illustrates the beam position (BP) X and the deviation. In FIG. 3, the horizontal axis indicates the coordinate along the circular path 172 of the accelerator 170, and the vertical axis indicates the beam position (BP) X for each BPM 181 to 184. The deviation ΔX 301 indicates the difference between the beam position (BP) X and the normal value indicating the designed beam orbit. For convenience, the upper side in FIG. 3 is the outside of the designed beam orbit as viewed from the center of the circular path 172 (the center of the accelerator 170). A range 302 not requiring adjustment is set with a certain width so as to sandwich the normal value indicating the designed beam orbit. If the BP (or deviation ΔX 301) in all the BPMs 181 to 184 falls within the range 302 not requiring adjustment, the parameters (parameter set) for controlling the accelerator do not need to be adjusted. On the other hand, when the BP in any of the BPMs 181-184 deviates from the adjustment-free range 302 and enters the adjustment-required range 303, a process for estimating the cause of the deviation and a process for adjusting (creating) the parameters (parameter set) for controlling the accelerator may be performed. Note that, when the BP in any of the BPMs 181-184 significantly deviates from the normal value and enters the alarm generation range 304, the information processing system 101 or the accelerator control device 240 may notify an operator or the like of an alarm, or may stop the particle beam therapy system 160 including the accelerator 170. The electromagnet information acquisition unit 212 acquires electromagnet information from the data accumulation server 250. The sensor data acquisition unit 213 acquires measurement values ​​of various sensors from the sensor data collection device 280. If the sensor data acquisition unit 213 can acquire measurement values ​​directly from the various sensors, the sensor data collection device 280 may not be required. The system information acquisition unit 214 acquires system information from the data accumulation server 250. The electromagnet information acquisition unit 212, the sensor data acquisition unit 213, and the system information acquisition unit 214 may be collectively referred to as the condition information acquisition unit 210. Furthermore, the various pieces of information acquired by the condition information acquisition unit 210 may be collectively referred to as condition information. The cause estimation unit 221, the parameter acquisition calculation unit 222, and the warm-up operation determination unit 223 may all operate when the deviation of the beam orbit 173 from the desired beam orbit (design beam orbit) deviates from the adjustment-free range. The cause estimation unit 221 estimates the cause of the deviation of the beam orbit 173 deviating from the adjustment-free range 302. The parameter acquisition calculation unit 222 acquires or calculates parameters (parameter set) for controlling the accelerator 170. The parameters here may be, for example, parameters for the amount of current I set in the electromagnets 191 to 194. The warm-up operation determination unit 223 determines whether or not a warm-up operation is performed and the mode of the warm-up operation, and determines parameters related to the warm-up operation. As will be shown later as a modified example, not only the warm-up operation but also the cooling operation may be determined. It is expected that a change in the temperature of the accelerator (for example, the temperature of an electromagnet used in the accelerator) will affect the beam orbit of the accelerator. To deal with this, the particle beam therapy system 160 is configured to be able to perform a warm-up operation or a cooling operation (including a temporary stop of the particle beam therapy system 160). The adjustment timing determination unit 224 determines the timing for adjusting parameters related to the control of the accelerator 170 in a future period (estimated period, subsequent period). The cause estimation unit 221, the parameter acquisition calculation unit 222, the warm-up operation determination unit 223, and the adjustment timing determination unit 224 may be collectively referred to as various calculation units 220. The display output control unit 231 controls some kind of display device or output device so as to display or output information showing the results output by the various calculation units 220 to the operator of the accelerator 170. The parameter setting unit 232 transfers and stores the parameters (parameter set) acquired by the parameter acquisition calculation unit 222 or the parameters (parameter set) further modified by the operator in response to an instruction from the operator of the accelerator 170 to the parameter storage device 241, thereby setting the parameters (parameter set) in the accelerator control device 240. Alternatively, the parameter setting unit 232 sets the parameters determined by the warm-up operation determination unit 223 or the parameters further modified by the operator in response to an instruction from the operator of the accelerator 170 to the accelerator control device 240. The parameter setting unit 232 may set the parameters (parameter set) acquired by the parameter acquisition calculation unit 222 or the parameters determined by the warm-up operation determination unit 223 in the accelerator control device 240 without waiting for an instruction from the operator of the accelerator 170.

[0016] The database (DB) 260 can provide the information processing system 101 with a group of records relating to the past operation history of the accelerator 170, machines of the same or similar type as the accelerator 170 that are installed in other locations, and synchrotrons in general. It is expected that the number of accelerators used in particle beam therapy systems will be greater than those used for other purposes. In other words, accelerators used in particle beam therapy systems are easy to accumulate records relating to their past operation history, so a well-equipped DB 260 can be realized. Here, the DB 260 may be built into the information processing system 101, may be connected to the information processing system 101 via an intranet, or may be remotely connected to the information processing system 101 via the Internet. Fig. 4 shows each of the components (items) included in each record of DB260. As shown in Fig. 4, each record 401 of DB260 may have 14 components (items). Assuming that the component number (item number) is represented by a number in brackets [ ], the following information may be included in each component (item) for each component number (item number). Note that the record of DB260 may include components (items) that do not exist among the following components (items), and may also include components (items) that are different from the following components (items). [1] The temperature of the outside air surrounding building 270. This information can be indicative of the season and time of day. [2] Room temperature in the room housing Accelerator-170. [3] Date or time associated with the information contained in this record. Of this information, the date information is information that suggests the season. Note that both date and time information may be present, or only date or only time information may be present, or time zone information may be present instead of time. [4] The time elapsed after the start-up or restart of the particle beam therapy system 160. For example, if the elapsed time is shorter than a predetermined time, information on the elapsed time is information suggesting the temperature of the accelerator 170 or the cooling water for the accelerator 170. [5] The temperature of the cooling water for the accelerator 170 (cooling water temperature). [6] Time elapsed since construction of Building 270. [7] The position of the floor surface on which the accelerator 170 is installed (the position of the accelerator installation surface). Although the causal relationship or correlation between this information and the operation of the accelerator 170 is not necessarily clear, there may be some causal relationship or correlation. [8] Parameters for controlling the accelerator 170 that were applied to each electromagnet (accelerator parameters before adjustment, e.g., current I or excitation B). [9] BP information measured per BPM (unadjusted BP), or deviation of BP from normal values.

[10] When BP or deviation is outside the adjustment-free range, this is a probable cause vector that indicates the cause of the deviation. (See Figure 5 below.)

[11] Whether any adjustments were made to the parameters controlling accelerator 170 on that date or time.

[12] When the parameters are adjusted, the parameters for controlling the accelerator 170 after adjustment are applied to each electromagnet.

[13] Whether or not the warm-up operation was performed on that date or time. (As will be shown later as a modified example, if a cooling operation can be performed, the component (item) of

[13] may include information on whether or not the cooling operation was performed on that date or time.)

[14] The remaining time of the warm-up operation, if a warm-up operation has been performed. It may also be the time period of the warm-up operation. (As will be shown later as a modified example, if a cooling operation can be performed, the component (item) of

[14] may include information on the remaining time of the cooling operation, if a cooling operation has been performed, or the time period of the cooling operation.) In the following, information on the components (items) shown in FIG. 4, or information corresponding to the components (items) (not limited to information contained in the records of DB 260), may be indicated by the numbers in brackets [ ].

[0017] The estimated cause vector, which is the tenth component of the record in DB 260, is also output by cause estimation unit 221 as an estimated result of the cause of the deviation of beam trajectory 173 (BP, deviation 301) from adjustment-free range 302. 5 shows a probable cause vector. The probable cause vector 501 may have seven components (items). The component numbers (item numbers) are represented by numbers with <>, and for each component number (item number), when the information of each component (item) is information indicating a set, it may have the following meaning. When the information of each component (item) is information indicating a reset, it may not have the following meaning. Note that the probable cause vector 501 may have a component (item) that does not exist among the following components (items), and may also have a component (item) that is different from the following components (items). <1> It is presumed that the cause was the temperature of the outside air (outdoor temperature) surrounding building 270. <2> It is believed that one of the causes was the room temperature in the room housing the accelerator 170. <3> The date or time is assumed to be included in the cause. <4> The time elapsed after the start-up or restart of the particle beam therapy system 160 is presumed to be one of the causes. <5> It is presumed that one of the causes is the temperature of the cooling water for the accelerator 170 (cooling water temperature). <6> It is presumed that the cause is the time that has passed since the construction of building 270. <7> It is presumed that the cause includes the position of the floor surface on which the accelerator 170 is installed (the position of the accelerator installation surface). In the following, information on the components (items) shown in FIG. 5 or information corresponding to the components (items) (not limited to information contained in the records of DB 260) may be indicated by numbers in brackets <>.

[0018] 3. Computer Architecture for Implementing the Embodiments of the Disclosure FIG. 6 illustrates a computer architecture for implementing an embodiment of the present disclosure. To realize the information processing system 101, an information processing device (e.g., a processor, a CPU) 601, a storage device (e.g., a memory) 602, a non-volatile recording medium (e.g., a non-volatile memory, a non-volatile disk device) 603, a drive for an external recording medium (e.g., a disk drive) 604, a display or output device (e.g., a display, a printer) 606, an input device (e.g., a mouse, a keyboard, an imaging device, a sensor) 607, a communication device (e.g., a communication device for wired communication, a communication device for wireless communication; this may be a network interface device (NIC) that controls communication with other systems, devices, or servers according to a predetermined protocol) 608, and some or all of an external input / output port 609 may be interconnected by an interconnection unit (e.g., a bus, a crossbar switch) 610. The non-volatile recording medium 603 may record a program 620a (for example, a beam monitoring program in the present disclosure) and various information. As the various information, for example, various tables 621 or various information 622 may be recorded in the non-volatile recording medium 603. Alternatively, a part or all of the above-mentioned program or various information may be acquired (accessed) from outside of FIG. 6. The external recording medium drive 604 can connect an external recording medium (for example, a portable recording disk (DVD, etc.), IC card, SD card) 605. Note that the program 620a (for example, a beam monitoring program in the present disclosure) and the above-mentioned various information may be transferred and stored from the external recording medium 605 to the non-volatile recording medium 603 or the storage device 602. In addition, the program 620a (e.g., the beam monitoring program in this disclosure) and the various information described above may be provided via the communication device 608, the external input / output port 609 or the input device 607 and stored in the non-volatile recording medium 603 or the storage device 602. In order for the architecture of FIG. 6 to function as the information processing system 101, each unit (functional unit) or part of each unit in the information processing system 101 (to execute one or a series of processes (steps)), the program 620a may be loaded into the storage device 602 (for example, from the non-volatile recording medium 603). The loaded program is indicated by 620b in FIG. 6. The information processing device 601 may then execute the program 620b (using various information present in the non-volatile recording medium 603, etc., as necessary). By executing the program 620b, the functions of the information processing system 101, each unit (functional unit) or part of each unit in the information processing system 101 are realized (one or a series of processes (steps) are executed). At this time, various buffers 623 temporarily formed in the storage device 602 may also be used as appropriate.

[0019] 4. Processing of the embodiments of the present disclosure In the following, among the processes of the embodiment of the present disclosure, the processes mainly performed by the cause estimation unit 221, the parameter acquisition calculation unit 222, and the warm-up operation determination unit 223 will be described. After that, the processes mainly performed by the adjustment timing determination unit 224 will be described. Note that it is not essential to perform all of the processes described below, and a beam monitoring method, beam monitoring program, or information processing system may be used to perform part of the processes described below.

[0020] 4.1. Cause estimation, parameter acquisition, and warm-up operation decision processing Among the processes that can be realized by the embodiments of the present disclosure, a series of processes that are performed when the BP (i.e., beam trajectory 173) in BPM 181-184 deviates from the adjustment-free range 302, such as estimating the cause of the deviation, adjusting (creating) parameters to bring the BP closer to the normal value (i.e., bring the beam trajectory 173 closer to the designed beam trajectory), or determining whether or not warm-up operation is occurring and its state, are described in Figures 7 to 24. Fig. 7 shows a main flowchart of the processes of cause estimation, parameter adjustment (creation), and warm-up operation determination. Figs. 8 to 12 are drawings for further explaining the cause estimation process of step 704 in Fig. 7. Figs. 13 to 18 are drawings for further explaining the parameter adjustment (creation) process of step 707 in Fig. 7. Fig. 19 is a drawing showing an example of a display screen by the display output process of step 707 or step 708 in Fig. 7. Figs. 20 to 24 are drawings for further explaining the warm-up operation determination process of step 709 in Fig. 7.

[0021] 4.1.1. Determination of deviation of beam orbit or beam path In step 701 of FIG. 7, the BP information acquisition unit 211 acquires BP information for each BPM 181-184 from the data accumulation server 250. The BP information is acquired by the BPMs 181-184 observing the beam 171. The BP information is temporarily stored in the log temporary storage device 242 from the BPMs 181-184. After that, the BP information is transferred from the log temporary storage device 242 to the data accumulation server 250 and accumulated. The BP information accumulated in the data accumulation server 250 is acquired by the BP information acquisition unit 211. Note that as long as the BP information acquisition unit 211 can acquire the BP information, there may be a modified example in which one or both of the log temporary storage device 242 and the data accumulation server 250 are not present and the number of hops until the BP information is acquired is small. This step 701 may be a beam position information acquisition step. Furthermore, the BP information may be the coordinates themselves in a coordinate system (e.g., a two-dimensional coordinate system or a one-dimensional coordinate system) that intersects with the traveling direction of the circular path 172, or may be a deviation 301 from a normal value indicating the designed beam trajectory, as shown in FIG. 3. In step 702 in Fig. 7, the BP information acquisition unit 211 judges whether the BP information acquired in step 701 has deviated from the adjustment-free range 302. For example, as shown in Fig. 3, the adjustment-free range 302 may be determined for the deviation 301, and the BP information acquisition unit 211 may judge whether the deviation 301 is included in the adjustment-free range 302 for each of the BPMs 181 to 184. Here, as shown in Fig. 3, the adjustment-necessary range 303 and the alarm occurrence range 304 may be determined in the range deviating from the adjustment-necessary range 302, and the BP information acquisition unit 211 may determine that the judgment result in step 702 in Fig. 7 is affirmative and that the processing from step 703 onward is to be performed if the deviation 301 is included in the adjustment-necessary range 303 for any of the BPMs 181 to 184 and the deviation 301 is not included in the alarm occurrence range 304 for any of the BPMs 181 to 184. Furthermore, when deviation 301 for any of BPMs 181 to 184 falls within alarm occurrence range 304, BP information acquisition unit 211 may notify an alarm to an operator of accelerator 170 or the like to urge them to stop accelerator 170. (In this case, information processing system 101 may not perform the processes from step 703 onward in FIG. 7, and accelerator control device 240 may perform processes to ensure the safety of accelerator 170.) Alternatively, alarm occurrence range 304 in FIG. 3 may not be set. In step 702, if the BP information acquisition unit 211 determines that the BP information is within the adjustment-free range 302 for all BPMs 181 to 184, control is returned to step 701, and BP information is again acquired for each BPM 181 to 184. In step 702, if the BP information acquisition unit 211 determines that the BP information is outside the adjustment-free range 302 for any of the BPMs 181 to 184, control is transferred to step 703. In step 703 of FIG. 7, information that is obtainable and is determined to be used in the processing in the following steps among the information on the BP, the electromagnet, the sensor, or the system is obtained. Here, when obtaining the information on the BP, the BP information obtaining unit 211 operates, when obtaining the information on the electromagnet (information on the amount of current I or the amount of excitation B, which is a parameter set in the electromagnets 191 to 194), the electromagnet information obtaining unit 212 operates, when obtaining the information on the sensor (measurement values ​​of the room temperature sensor 281, the cooling water temperature sensor 282, the installation surface position sensor 283, or the outside air temperature sensor 284), the sensor data obtaining unit 213 operates, and when obtaining the information on the system (date or time, the elapsed time after the start-up or restart of the particle beam therapy system 160, or the elapsed time after the building 270 was constructed), the system information obtaining unit 214 operates. Here, as shown in FIG. 2, regarding the information on the BP, the electromagnet, or the system, each piece of information may be obtained via the log temporary storage device 242 and the data accumulation server 250. 2, the sensor information may be acquired via the sensor data collecting device 280. Alternatively, each piece of information may be acquired without going through any of the log temporary storage device 242, the data accumulation server 250, or the sensor data collecting device 280 in FIG. In step 703, after acquiring information that is determined to be acquireable and to be used in the processing in the subsequent steps from among the information on the BP, electromagnet, sensor, or system, the information processing system 101 can transfer control to the step group of steps 704 to 708 (processing of cause estimation and parameter adjustment (creation)) or the step group of steps 709 to 711 (processing of warm-up operation determination). The information processing system 101 may sequentially execute the process of cause estimation and parameter adjustment (creation) and the process of warm-up operation determination, or may execute the process of cause estimation and parameter adjustment (creation) and the process of warm-up operation determination simultaneously in parallel. This step 703 may be a beam position information acquisition step. Also, this step 703 may be a condition information acquisition step. Based on the condition information acquired in this condition information acquisition step and the BP information for each BPM acquired in the beam position information acquisition step, parameters to be set for each electromagnet are acquired in the subsequent step 706 (parameter acquisition step), and appropriate values ​​are set as the parameters to be set for each electromagnet. Furthermore, based on the condition information acquired in the condition information acquisition step and the BP information for each BPM acquired in the beam position information acquisition step, the subsequent step 709 (warm-up operation determination step) determines whether or not warm-up operation is being performed and the nature of the warm-up operation (for example, the remaining time of warm-up operation (the duration of warm-up operation from the current point in time)), and appropriate parameters for the warm-up operation are set.

[0022] 4.1.2. Estimation of the cause of beam orbit or BP deviation (A) 7, the cause estimation unit 221 estimates the cause of the deviation of the BP information from the adjustment-free range 302 in any of the BPMs 181 to 184. The cause estimation unit 221 performs the process of cause estimation using each piece of information obtained in step 703. Three methods will be described below as examples of the process of cause estimation executed by the cause estimation unit 221. The three methods are a rule-based method, a DB search method, and a cause estimation model method. The cause estimation unit 221 may execute any one of these three methods. Alternatively, the cause estimation unit 221 may selectively execute any one of these three methods depending on whether a predetermined condition is satisfied. Alternatively, the cause estimation unit 221 may ensemble (for example, take a majority vote) the cause estimation results obtained by each of these three methods to obtain a final cause estimation result. Furthermore, the cause estimation unit 221 may perform cause estimation using a method other than these three methods. This step 704 may be a cause estimation step. This cause estimation step estimates the cause of the deviation of the beam orbit 173 (BP information) of the circulating beam 171 from the adjustment-free range 302, so that the information processing system 101 can provide an operator or the like with information for making a decision when controlling the accelerator 170.

[0023] 4.1.2.1. Rule-based cause estimation (A-1) Fig. 8 shows a flowchart of rule-based cause estimation processing that may be performed by the cause estimation unit 221 in step 704 in Fig. 7. In the flowchart in Fig. 8, the cause estimation unit 221 may obtain a cause estimation result as an estimated cause vector in the format shown in Fig. 5. In step 801 of FIG. 8, the cause estimation unit 221 calculates all components (items) (e.g., <1> ~ <7> ) is the reset value. 8, the cause estimation unit 221 checks the positional relationship between the designed beam orbit and the beam orbit 173 for the information of the BP acquired in step 701 and determined to deviate from the adjustment-free range 302 in step 702, and judges whether the beam orbit 173 (or the deviation 301) is located on the same side as the designed beam orbit (or the normal value) in any of the BPMs 181 to 184. In other words, the cause estimation unit 221 judges whether the situation is such that the beam orbit 173 is deviated outward from the designed beam orbit (as viewed from the center of the accelerator 170) in any of the BPMs 181 to 184, or the beam orbit 173 is deviated inward from the designed beam orbit (as viewed from the center of the accelerator 170) in any of the BPMs 181 to 184. If the determination result in step 802 is positive, the cause estimation unit 221 transfers control to step 803. If the determination result in step 802 is positive, that is, in any of BPMs 181 to 184, the beam trajectory 173 deviates outward from the designed beam trajectory (as viewed from the center of the accelerator 170), or in any of BPMs 181 to 184, the beam trajectory 173 deviates inward from the designed beam trajectory (as viewed from the center of the accelerator 170), then, based on experience, it is often safe to estimate that the cause of the deviation of the beam trajectory 173 is derived from some temperature related to the accelerator 170. Of the components (items) of the estimated cause vector in the format shown in FIG. 5, <1> and <2> and <3> and <4> and <5> The above components (items) may be responsible for some temperature-related issues associated with the accelerator 170. On the other hand, if the determination result in step 802 is negative, the cause estimation unit 221 transfers control to step 813. If the determination result in step 802 is negative, the deviation (distortion) of the beam trajectory 173 from the designed beam trajectory becomes complicated, and it is empirically estimated that the deviation of the beam trajectory 173 is caused by something other than the temperature of the accelerator 170. This step 802 may be a trajectory comparison step. When it is expected that the adjustment (creation) of a parameter (set) for correcting the deviation of the beam trajectory 173 from the design beam trajectory can be automated to a certain extent by such a trajectory comparison step, the adjustment (creation) process of the parameter (set) can be performed in the parameter acquisition step shown in step 706. 1 and 2, it may be possible to determine whether the beam trajectory 173 (or deviation 301) is located on the same side of the designed beam trajectory (or normal value) only for each of the BPMs included in a pair of BPMs at opposing positions (for example, a pair of BPMs 181 and 183).Simplifying the process in this way can reduce the processing load of step 802, for example, when the accelerator has a large number of BPMs. 8, the cause estimation unit 221 judges whether the outside air temperature information (measurement value of the outside air temperature sensor 284) acquired in step 703 is characteristic. For example, in the case where the accelerator 170 is assumed to be operated in a standard climate, if the outside air temperature information is higher than a predetermined first (high temperature) threshold value or if the outside air temperature information is lower than a predetermined second (low temperature) threshold value, the cause estimation unit 221 judges that the outside air temperature information is characteristic. Step 804 is executed only when the outside air temperature information is judged to be characteristic. In step 804 of FIG. 8, the cause estimation unit 221 calculates the estimated cause vector in the format shown in FIG. <1> The components (items) of are set to the values. 8, the cause estimation unit 221 judges whether the room temperature information acquired in step 703 (the measured value of the room temperature sensor 281) is characteristic. For example, in the case where the accelerator 170 is assumed to be operated at a standard room temperature, if the room temperature information is higher than a predetermined third (high temperature) threshold or if the room temperature information is lower than a predetermined fourth (low temperature) threshold, the cause estimation unit 221 judges that the room temperature information is characteristic. Step 806 is executed only when the room temperature information is judged to be characteristic. In step 806 of FIG. 8, the cause estimation unit 221 calculates the estimated cause vector in the format shown in FIG. <2> The components (items) of are set to the values. 8, the cause estimation unit 221 judges whether the date or time information acquired in step 703 (the date or time information linked to the BP information, the electromagnet information, and the sensor information) is characteristic. For example, in the case where the accelerator 170 is assumed to be operated in a standard climate or time zone, if the date information belongs to summer or winter, or if the time information indicates several hours around noon in summer or nighttime in winter, the cause estimation unit 221 judges that the date or time information is characteristic. Only when the date or time information is judged to be characteristic, step 808 is executed. In step 808 of FIG. 8, the cause estimation unit 221 calculates the estimated cause vector in the format shown in FIG. <3> The components (items) of are set to the values. 8, the cause estimation unit 221 judges whether the time elapsed after the start-up or restart of the particle therapy system 160 acquired in step 703 is characteristic. Empirically, it is estimated that the time elapsed after the start-up or restart of the particle therapy system 160 is related to the temperature of each component of the accelerator 170 including the cooling water. For example, it is estimated that the temperature of each component of the accelerator 170, including the cooling water, remains low immediately after the start-up or restart of the particle therapy system 160. Therefore, if the information on the elapsed time after the start-up or restart of the particle therapy system 160 indicates a time immediately after the start-up or restart of the particle therapy system 160 (the elapsed time is smaller than a predetermined fifth (time) threshold), the cause estimation unit 221 determines that the information on the elapsed time after the start-up or restart of the particle therapy system 160 is characteristic. On the other hand, when the time elapsed after the start-up or restart of the particle beam therapy system 160 is too long, the temperature of each component of the accelerator 170 including the cooling water may become higher than expected. Therefore, if the information on the time elapsed after the start-up or restart of the particle beam therapy system 160 indicates that the time elapsed is too long (the time elapsed is greater than a predetermined sixth (time) threshold), the cause estimation unit 221 may determine that the information on the time elapsed after the start-up or restart of the particle beam therapy system 160 is characteristic. Step 810 is executed only if the information on the elapsed time after the start-up or restart of the particle therapy system 160 is determined to be characteristic. In step 810 of FIG. 8, the cause estimation unit 221 calculates the estimated cause vector in the format shown in FIG. <4> The components (items) of the estimated cause vector are set to the values. <4> The component (item) may be configured to be capable of distinguishing between a time elapsed after startup or restart of the particle beam therapy system 160 being less than a fifth (time) threshold and a time elapsed after startup or restart of the particle beam therapy system 160 being greater than a sixth (time) threshold. 8, the cause estimation unit 221 judges whether the coolant temperature information (the measured value of the coolant temperature sensor 282) acquired in step 703 is characteristic. For example, in a case where the accelerator 170 is assumed to be operated using coolant at a standard temperature, if the coolant temperature information is lower than a predetermined seventh (low temperature) threshold, the cause estimation unit 221 judges that the coolant temperature information is characteristic. Incidentally, when the coolant time information is higher than the predetermined eighth (high temperature) threshold, the cause estimating unit 221 may also determine that the coolant temperature information is characteristic. Step 812 is executed only if the coolant temperature information is determined to be characteristic. In step 812 of FIG. 8, the cause estimation unit 221 calculates the estimated cause vector in the format shown in FIG. <5> The components (items) of the estimated cause vector are set to the values. <5> The component (item) may be capable of distinguishing between coolant temperature information being lower than a predetermined seventh (low temperature) threshold and coolant temperature information being higher than a predetermined eighth (high temperature) threshold. 8, the cause estimation unit 221 judges whether the time elapsed since the building 270 was constructed, acquired in step 703, is characteristic. Although the causal relationship or correlation between the time elapsed since the building 270 was constructed and the operation of the accelerator 170 is not necessarily clear, for example, if the time elapsed since the building 270 was constructed is longer than a predetermined ninth (time) threshold, the cause estimation unit 221 may judge that the time elapsed since the building 270 was constructed is characteristic. Step 814 is executed only when it is judged that the time elapsed since the building 270 was constructed is characteristic. In step 814 of FIG. 8, the cause estimation unit 221 calculates the estimated cause vector in the format shown in FIG. <6> The components (items) of are set to the values. In step 815 of Fig. 8, the cause estimation unit 221 judges whether the position of the surface of the floor on which the accelerator 170 is installed (the measurement value by the installation surface position sensor 283) acquired in step 703 is characteristic. Although the causal relationship or correlation between the position of the surface of the floor on which the accelerator 170 is installed and the operation of the accelerator 170 is not necessarily clear, for example, if the deviation of the plane of the floor on which the accelerator 170 is installed from the plane perpendicular to the gravity vector due to the geographical characteristics of the place where the accelerator 170 is installed or crustal movement due to an earthquake is greater than a predetermined tenth (deviation) threshold, the cause estimation unit 221 may judge that the position of the surface of the floor on which the accelerator 170 is installed is characteristic. Only when it is judged that the position of the surface of the floor on which the accelerator 170 is installed is characteristic, step 816 is executed. In step 816 of FIG. 8, the cause estimation unit 221 calculates the estimated cause vector in the format shown in FIG. <7> The components (items) of are set to the values. If rules that serve as criteria for judgment are determined in advance, the cause inference process based on the rules as shown in FIG. 8 can be easily realized.

[0024] 4.1.2.2. Estimation of the cause by DB search (A-2) Fig. 9 shows a flowchart of a cause estimation process by DB search that can be performed by the cause estimation unit 221 in step 704 in Fig. 7. In the flowchart in Fig. 8, the cause estimation unit 221 obtains a cause estimation result in the form of an estimated cause vector shown in Fig. 5. In step 901 in Fig. 9, the cause estimation unit 221 compares the component (item) information in each record 401 in the format shown in Fig. 4 from the database (DB) 260 with the component (item) information obtained from the information on the BP, electromagnet, sensor, and system acquired in step 701 or step 703. The components (items) compared here may be, for example, some or all of the nine components (items) consisting of [1] to [9] among the components (items) shown in Fig. 4. Furthermore, the cause estimation unit 221 may perform a comparison process on components (items) other than the components (items) listed above. 9, the cause estimation unit 221 selects a record having many similarities with each piece of information on the BP, electromagnet, sensor, and system (information associated with the deviation of the beam trajectory 173 (BP, deviation 301) from the adjustment-free range 302) acquired in step 701 or step 703 from among the records 401 compared in step 901. For example, the cause estimation unit 221 may obtain a score representing the similarity for each record by scoring the difference obtained in the comparison process in step 901 for each of the nine components (items) described above, and select the record having the highest score in step 902. 9, the cause estimation unit 221 may extract information (estimated cause vector in the format shown in FIG. 5) of the component (item)

[10] , which is the component (item) included in the record selected in step 902, and treat the extracted estimated cause vector as an estimated cause vector corresponding to each piece of information on the BP, electromagnet, sensor, and system (information associated with the deviation of the beam trajectory 173 from the adjustment-free range 302) acquired in step 701 or step 703. This estimated cause vector becomes the result of the cause estimation in FIG. 9. The above-described cause inference process using a database search can relatively easily realize cause inference using the accelerator's past operation history information.

[0025] 4.1.2.3. Estimation of the cause using a cause estimation model (A-3) Fig. 10 shows inputs and outputs of the cause estimation model 1000 when the cause estimation unit 221 executes the cause estimation model 1000 to perform the cause estimation process in step 704 in Fig. 7. The cause estimation unit 221 realizes the cause estimation model 1000 based on a program and a group of model parameters for the cause estimation model 1000. Note that when realizing the cause estimation model 1000, a part or the whole of it may be realized by hardware rather than by program control. The cause estimation model 1000 receives as input the BP, electromagnet, sensor, and system information (information linked to the deviation of the beam trajectory 173 from the adjustment-free range 302) acquired in step 701 or step 703, and outputs an estimated cause vector. In FIG. 10, possible inputs to the cause estimation model 1000 include BP information (information corresponding to [9] in FIG. 4), electromagnet information (information corresponding to [8] in FIG. 4), sensor information (information corresponding to [1], [2], [5], and [7] in FIG. 4), and system information (information corresponding to [3], [4], and [6] in FIG. 4). The output of the cause estimation model 1000 is an estimated cause vector (information corresponding to

[10] in FIG. 4, or information in the format shown in FIG. 5 ( <1> ~ <7> There may be information about 11 shows a flowchart of machine learning for training model parameters to construct the cause inference model 1000. This machine learning may be executed by the information processing system 101 or may be executed by some system other than the information processing system 101. In step 1101 in FIG. 11, the system that performs machine learning reads one record from a record group to be used for machine learning, from among the records held by the DB 260 (records in the format shown in FIG. 4). In step 1102 of FIG. 11, the system performing machine learning inputs some or all of the nine components (items) consisting of [1] to [9] of the components (items) contained in the record read in step 1101 to the cause estimation model 1000. In step 1103 of FIG. 11, the system that executes machine learning obtains an output from the cause estimation model 1000 according to the input of each component (item) to the cause estimation model 1000 performed in step 1102. Here, the output from the cause estimation model 1000 is an estimated cause vector (information corresponding to

[10] in FIG. 4, or information in the format shown in FIG. 5 ( <1> ~ <7> The format is (information about the In step 1104 of FIG. 11, the system performing machine learning uses the information (presumed cause vector) of the component (item)

[10] among the components (items) of the record read in step 1101 as training data. 11, the system that performs machine learning treats the difference between the estimated cause vector obtained in step 1103 and the estimated cause vector that is the teacher data obtained in step 1104 as an error. Then, the system that performs machine learning adjusts (trains) the model parameters of the cause estimation model 1000 by an error backpropagation method or the like using the error. In step 1106 of Fig. 11, the system that performs machine learning determines whether all records included in the record group to be used for machine learning have been read out from among the records held by the DB 260. If the result of the determination indicates that there are records remaining that have not yet been read out in step 1101 as records to be used for machine learning, the system that performs machine learning returns control to step 1101, determines the next record to be read out from among the records that have not yet been read out in step 1101 as records to be used for machine learning, and reads out that record. If the result of the determination indicates that all records to be used for machine learning have been read out in step 1101, the machine learning process of the cause estimation model 1000 is completed. FIG. 12 shows a flowchart of the cause estimation process in the case where the cause estimation unit 221 executes the cause estimation model 1000 when performing the cause estimation process in step 704 in FIG. In step 1201 of FIG. 12, the cause estimation unit 221 inputs information corresponding to some or all of the nine components (items) consisting of [1] to [9] (represented by the components (items) shown in FIG. 4) from the information on BP, electromagnets, sensors, and the system acquired in step 701 or step 703 (information linked to the deviation of the beam trajectory 173 (BP, deviation 301) from the adjustment-free range 302) into the cause estimation model 1000. In step 1202 of FIG. 12, the cause estimation unit 221 outputs an estimated cause vector (information corresponding to

[10] in FIG. 4 or information in the format shown in FIG. 5 ( <1> ~ <7> This estimated cause vector is the result of the cause estimation in FIG. The cause inference process using a model trained by machine learning as described above can realize cause inference that fully reflects the accelerator's past operating history information.

[0026] 4.1.2.4. Judgment on the feasibility of parameter adjustment (creation) processing based on the results of cause estimation processing In step 704 in Fig. 7, the cause estimation unit 221 estimates the cause of the deviation of the BP information from the adjustment-free range 302 in any of the BPMs 181 to 184. In step 705 after that, the cause estimation unit 221 judges whether or not any of the estimated causes obtained as a result of the cause estimation is a cause derived from some temperature related to the accelerator 170. For example, among the components (items) of the estimated cause vector in the format shown in Fig. 5, <1> ~ <5> These five components (items) may be treated as temperature-related causes. If the result of this determination is positive (if any of the estimated causes is some temperature-related cause related to the accelerator 170), the information processing system 101 transfers control to step 706. If the result of the determination is negative, the information processing system 101 transfers control to step 708. In step 705, the cause estimation unit 221 may perform the same judgment process as step 802 in Fig. 8. That is, in step 705, the cause estimation unit 221 may check the positional relationship between the designed beam orbit and the beam orbit 173 for the information of the BP acquired in step 701 and judged to deviate from the adjustment-free range 302 in step 702, and judge whether the beam orbit 173 (or the deviation 301) is located on the same side as the designed beam orbit (or the normal value) in any of the BPMs 181 to 184. That is, the cause estimation unit 221 may judge whether the situation is such that the beam orbit 173 is deviated outward from the designed beam orbit (as viewed from the center of the accelerator 170) in any of the BPMs 181 to 184, or the beam orbit 173 is deviated inward from the designed beam orbit (as viewed from the center of the accelerator 170) in any of the BPMs 181 to 184. If the determination result in step 705 is positive, that is, if the beam trajectory 173 deviates outward from the designed beam trajectory (as viewed from the center of the accelerator 170) in any of the BPMs 181 to 184, or if the beam trajectory 173 deviates inward from the designed beam trajectory (as viewed from the center of the accelerator 170) in any of the BPMs 181 to 184, the information processing system 101 may shift control to step 706. On the other hand, if the determination result in step 705 is negative, the information processing system 101 may shift control to step 708. This step 705 may be a trajectory comparison step. When it is expected that the adjustment (creation) of a parameter (set) for correcting the deviation of the beam trajectory 173 from the design beam trajectory can be automated to a certain extent by such a trajectory comparison step, the adjustment (creation) process of the parameter (set) can be performed in the parameter acquisition step shown in step 706 (and the parameter display output step shown in step 707 can be performed). Alternatively, in step 705, the cause estimation unit 221 may perform a process that simplifies the judgment process in step 802 in Fig. 8. Specifically, in step 705, the cause estimation unit 221 may determine whether the beam trajectory 173 (or the deviation 301) is located on the same side as viewed from the designed beam trajectory (or the normal value) in any BPM included in a pair of BPMs (for example, a pair of BPMs 181 and 183) that are located opposite each other among the BPMs 181 to 184. In this case, the BPM 181 may be the first beam monitor, and the BPM 183 may be the second beam monitor. Also, the electromagnet 194 may be the first bending electromagnet, the electromagnet 191 may be the second bending electromagnet, the electromagnet 192 may be the third bending electromagnet, and the electromagnet 193 may be the fourth bending electromagnet. By simplifying the process in this way, it is possible to reduce the processing load of step 802 when, for example, the accelerator has a large number of BPMs. If the determination result of step 705 is positive, the deviation 301 has the same sign (uniformly positive or uniformly negative) in all the BPMs 181-184 (or in all the BPMs 181 and 183 at corresponding positions), and the degree of complexity of the deviation of the beam trajectory 173 from the designed beam trajectory tends to be low. In such a case, for example, when a bending electromagnet also having a steering function for adjusting the magnetic field to adjust the beam trajectory 173 is used as the electromagnets 191-194, most of the deviation of the beam trajectory 173 can be corrected by adjusting the bending component of each of the components of the magnetic field generated by the bending electromagnet (bending component and steering component). Here, since the bending component is easier to adjust than the steering component, the process of adjusting the parameters of the electromagnets is easier to realize. In addition, since there are many cases in which the judgment result in step 705 is positive, records of the cases are likely to be accumulated as past operation history. Therefore, for cases in which the judgment result in step 705 is positive, it can be said that performing the parameter set adjustment (creation) process in step 706 is quite easy to realize, for example, by directly utilizing the database (DB) 260 that accumulates the records of the cases, or by using a model trained by machine learning using the records of the DB 260.

[0027] 4.1.3. Adjusting (creating) parameter sets (B) 7, the parameter acquisition calculation unit 222 adjusts (creates) a set of parameters (for example, parameters of the current amount I or excitation amount B used to control each of the electromagnets 191-194) used to control the accelerator 170 in order to bring the BP information in each of the BPMs 181-184 closer to a normal value (bringing the absolute value of the deviation 301 closer to zero, bringing the beam trajectory 173 closer to the designed beam trajectory). Alternatively, the parameter acquisition calculation unit 222 calculates a set of parameters. Three methods will be described below as examples of the process of adjusting (creating) a parameter set executed by the parameter acquisition calculation unit 222. The three methods are a method using Closed Orbit Distortion (COD) correction, a method using a DB search, and a method using a parameter acquisition model. The parameter acquisition calculation unit 222 may execute any one of these three methods. Alternatively, the parameter acquisition calculation unit 222 may selectively execute any one of these three methods depending on whether a predetermined condition is satisfied. For example, in a case where the parameter acquisition calculation unit 222 is capable of executing the COD correction method and the DB search method, if a record useful for adjusting (creating) a parameter set is found in the DB 260, the parameter acquisition calculation unit 222 may adjust (create) the parameter set using the DB search method, and if a useful record is not found, the parameter set may be adjusted (created) using the COD correction method. Alternatively, if the deviation 301 to be corrected is relatively small, the COD correction method may be prioritized. Alternatively, the parameter acquisition calculation unit 222 may ensemble (for example, take an average) the results of adjusting (creating) the parameter set obtained by each of these three methods to obtain a final cause estimation result. Furthermore, the parameter acquisition calculation unit 222 may adjust (create) the parameter set using a method other than these three methods. This step 706 may be a parameter acquisition step. By such a parameter acquisition step, it is possible to automate the adjustment (creation) of a parameter (set) for making the beam trajectory 173 in the accelerator 170 closer to the designed beam trajectory. In the following, after explaining the three methods, we will also explain a method for creating a parameter set when the parameter set is represented as a three-dimensional array (a method that can be used in common for the three methods above).

[0028] 4.1.3.1. Adjustment (creation) of parameter set by COD correction (B-1) FIG. 13 shows a flowchart of a process of adjusting (creating) a parameter set by COD correction, which can be performed by the parameter acquisition calculation unit 222 in step 706 in FIG. 13, the parameter acquisition calculation unit 222 acquires (calculates) a deviation ΔX of the BP from a normal value indicating a design beam trajectory as information on the BP for each BPM. This deviation ΔX is shown as 301 in FIG. 13, the parameter acquisition calculation unit 222 acquires an adjustment amount ΔB of the excitation amount for each of the electromagnets 191 to 194 from the deviation ΔX for each of the BPMs 181 to 184 based on the formula for COD correction. When the beam positions (BP) for each of the BPMs 181 to 184 are X_181, X_182, X_183, and X_184, and the excitation amounts for each of the electromagnets 191 to 194 are B_191, B_192, B_193, and B_194, the vector of the beam position (BP) X [X_181, X_182, X_183, and X_184] and the vector of the excitation amount B [B_191, B_192, B_193, and B_194] are related by some function F. In other words, it can be expressed as [X_181, X_182, X_183, X_184] = F ([B_191, B_192, B_193, B_194]). Therefore, the parameter acquisition calculation unit 222 can calculate the adjustment amount ΔB of the excitation amount for each of the electromagnets 191-194 from the deviation ΔX for each of the BPMs 181-184 by using the function F (or the inverse function of the function F). Alternatively, when ΔX and ΔB can be regarded as minute quantities, a linear mapping relationship may be established between the deviation vector [ΔX_181, ΔX_182, ΔX_183, ΔX_184] related to [X_181, X_182, X_183, X_184] and the adjustment amount vector [ΔB_191, ΔB_192, ΔB_193, ΔB_194] related to [B_191, B_192, B_193, B_194]. In other words, when a matrix showing a linear mapping is P, the deviation vector (or the deviation vector multiplied by (-1)) may be expressed by the product of the matrix P and the adjustment amount vector. In that case, the parameter acquisition calculation unit 222 can calculate the adjustment amount ΔB of the excitation amount for each of the electromagnets 191 to 194 from the deviation ΔX for each of the BPMs 181 to 184 by using the matrix P (or the inverse matrix of the matrix P). 13, the parameter acquisition calculation unit 222 acquires (calculates) an adjustment amount ΔI of the amount of current for each of the electromagnets 191-194 from the adjustment amount ΔB of the amount of excitation for each of the electromagnets 191-194 obtained in step 1302. Note that the parameter acquisition calculation unit 222 can use a relationship (relational expression) between the amount of current I and the amount of excitation B for each of the electromagnets 191-194 that has been obtained in advance. 13, the parameter acquisition calculation unit 222 creates an adjusted parameter set from the adjustment amount ΔI of the current amount for each electromagnet obtained in step 1303. For example, the parameter set is a three-dimensional array of parameters. The creation of a parameter set that is a three-dimensional array will be explained later in "4.1.3.4. Creation of a parameter set as a three-dimensional array." The above-described adjustment (creation) process of the parameters (set) through COD correction can realize the adjustment (creation) of the parameters (set) even in a situation where there is little accumulated operation history information.

[0029] 4.1.3.2. Adjustment (creation) of parameter set by DB search (B-2) FIG. 14 shows a flowchart of a parameter set adjustment (creation) process by DB search that can be performed by the parameter acquisition calculation unit 222 in step 706 in FIG. In step 1401 in Fig. 14, the parameter acquisition calculation unit 222 compares the component (item) information in each record 401 in the format shown in Fig. 4 from the database (DB) 260 with the component (item) information obtained from each piece of information on the BP, electromagnet, sensor, and system acquired in step 701 or step 703. The components (items) compared here may be, for example, information corresponding to some or all of the nine components (items) consisting of [1] to [9] among the components (items) shown in Fig. 4. Furthermore, the parameter acquisition calculation unit 222 may perform a comparison process on components (items) other than the components (items) listed above. 14, the parameter acquisition calculation unit 222 selects a record having many similarities with each piece of information on the BP, electromagnet, sensor, and system (information associated with the deviation of the beam trajectory 173 (BP, deviation 301) from the adjustment-free range 302) acquired in step 701 or step 703 from among the records 401 compared in step 1401. For example, the parameter acquisition calculation unit 222 may obtain a score for the similarity of each record by scoring the difference obtained in the comparison process in step 1401 for each of the nine components (items) described above, and select the record having the highest score in step 1402. In step 1403 of Fig. 14, the parameter acquisition calculation unit 222 creates an adjusted parameter set based on the information of the component (item)

[12] included in the record selected in step 1402. For example, the parameter acquisition calculation unit 222 may set the difference between the parameter indicated by the component (item)

[12] included in the selected record and the parameter as the electromagnet information acquired in step 703 as the adjustment amount ΔI of the current amount for each electromagnet, and then create an adjusted parameter set from this adjustment amount ΔI of the current amount for each electromagnet. For example, the parameter set is a three-dimensional array of parameters. The creation of a parameter set that is a three-dimensional array will be described later in "4.1.3.4. Creation of a parameter set as a three-dimensional array". The above-described parameter set adjustment (creation) process by DB search can relatively easily realize the adjustment (creation) of a parameter set using the accelerator's past operation history information.

[0030] 4.1.3.3. Adjustment (creation) of parameter set using parameter acquisition model (B-3) Fig. 15 shows inputs and outputs of the parameter acquisition model 1500 when the parameter acquisition calculation unit 222 executes the parameter acquisition model 1500 to adjust (create) the parameter set in step 706 in Fig. 7. The parameter acquisition calculation unit 222 realizes the parameter acquisition model 1500 based on a program and a group of model parameters for the parameter acquisition model 1500. Note that when realizing the parameter acquisition model 1500, a part or the whole may be realized by hardware rather than by program control. The parameter acquisition model 1500 receives as input the information of the BP, electromagnet, sensor, and system acquired in step 701 or step 703 (information associated with the deviation of the beam trajectory 173 (BP, deviation 301) from the adjustment-free range 302), and outputs adjusted parameters (electromagnet information). In FIG. 15, the inputs of the parameter acquisition model 1500 may include BP information (information corresponding to [9] in FIG. 4), electromagnet information (information corresponding to [8] in FIG. 4), sensor information (information corresponding to [1], [2], [5], and [7] in FIG. 4), and system information (information corresponding to [3], [4], and [6] in FIG. 4). The output of the parameter acquisition model 1500 may include adjusted parameters (electromagnet information) (information corresponding to

[12] in FIG. 4). 16 shows a flowchart of machine learning for training model parameters to construct the parameter acquisition model 1500. This machine learning may be executed by the information processing system 101 or may be executed by some system other than the information processing system 101. In step 1601 in FIG. 16, the system that performs machine learning reads one record from a set of records to be used for machine learning, among the records held by the DB 260 (records in the format shown in FIG. 4). In step 1602 of FIG. 16, the system performing machine learning inputs some or all of the nine components (items) consisting of [1] to [9] of the components (items) contained in the record read in step 1601 to the parameter acquisition model 1500. 16, the system that executes machine learning obtains an output from the parameter acquisition model 1500 according to the input of each component (item) to the parameter acquisition model 1500 performed in step 1602. Here, the output from the parameter acquisition model 1500 is in the form of adjusted parameters (electromagnet information) (information corresponding to

[12] in FIG. 4). In step 1604 of FIG. 16, the system that performs machine learning sets the information of the component (item)

[12] among the components (items) of the record read in step 1601 as training data. 16, the system that performs machine learning treats the difference between the adjusted parameters (electromagnet information) obtained in step 1603 and the adjusted parameters (electromagnet information) that are the teacher data obtained in step 1604 as an error. Then, the system that performs machine learning adjusts (trains) the model parameters of the parameter acquisition model 1500 by an error backpropagation method or the like using the error. 16, the system that performs machine learning determines whether all records included in the record group to be used for machine learning have been read out from among the records held by DB 260. If the result of the determination indicates that there are records remaining that have not yet been read out in step 1601 as records to be used for machine learning, the system that performs machine learning returns control to step 1601, determines the next record to be read out from among the records that have not yet been read out in step 1601 as records to be used for machine learning, and reads out that record. If the result of the determination indicates that all records to be used for machine learning have been read out in step 1601, the machine learning process of the parameter acquisition model 1500 is completed. FIG. 17 shows a flowchart of the parameter set adjustment (acquisition) process in the case where the parameter acquisition calculation unit 222 executes the parameter set adjustment (creation) process in step 704 in FIG. In step 1701 of FIG. 17, the parameter acquisition calculation unit 222 inputs information corresponding to some or all of the nine components (items) consisting of [1] to [9] from the information on BP, electromagnet, sensor, and system acquired in step 701 or step 703 (information linked to the deviation of the beam trajectory 173 (BP, deviation 301) from the adjustment-free range 302) into the parameter acquisition model 1500. In step 1702 in FIG. 17, the parameter acquisition calculation unit 222 acquires, as an output from the parameter acquisition model 1500, adjusted parameters (electromagnet information for each electromagnet) (information corresponding to

[12] in FIG. 4). In step 1703 of FIG. 17, the parameter acquisition calculation unit 222 creates an adjusted parameter set based on the adjusted parameters (electromagnet information for each electromagnet) (information corresponding to

[12] in FIG. 4) obtained from the parameter acquisition model 1500 in step 1702. For example, the parameter acquisition calculation unit 222 may set the difference between the adjusted parameters (electromagnet information for each electromagnet) (information corresponding to

[12] in FIG. 4) obtained from the parameter acquisition model 1500 and the parameters as the electromagnet information obtained in step 703 as an adjustment amount ΔI of the current amount for each electromagnet, and create an adjusted parameter set from this adjustment amount ΔI of the current amount for each electromagnet. For example, the parameter set is a three-dimensional array of parameters. The creation of a parameter set that is a three-dimensional array will be described later in "4. Creation of a parameter set as a three-dimensional array in 3.4.1 of 4". The parameter set adjustment (creation) process using a model trained by machine learning as described above can adjust (create) a parameter set that fully reflects the accelerator's past operating history information.

[0031] 1.3.4 Creating a Parameter Set as a Three-Dimensional Array In step 1304 in FIG. 13, step 1403 in FIG. 14, and step 1703 in FIG. 17, when the parameter acquisition calculation unit 222 creates a parameter set which is a three-dimensional array from the adjustment amount ΔI of the current amount for each electromagnet, the parameter acquisition calculation unit 222 may execute the processing of the flowchart shown in FIG. 18. The parameter set, which is a three-dimensional array, has as its array indexes a target energy amount V, a time T, and an electromagnet position C. In other words, the target value of the amount of energy that the beam 171 obtains in the accelerator is set to V, and the parameter of the amount of current I to be applied to the electromagnet at the electromagnet position C at time T is set to one element I(V, T, C) of the three-dimensional array. Fig. 18 is a flowchart for the parameter acquisition calculation unit 222 to sequentially obtain the elements I(V,T,C) of the parameter set, which is a three-dimensional array as described above. To briefly explain the process in Fig. 18, I(V,T,C) is sequentially obtained by a triple loop process using V, T, and C as indexes, respectively. 18, the parameter acquisition calculation unit 222 initializes the indexes of the elements of the three-dimensional array to be recalculated based on the adjustment amount ΔI of the current amount for each electromagnet. For example, if the range of the values ​​of the indexes is from V_S to V_E for the target energy amount V, from T_S to T_E for the time T, and from C_S to C_E for the electromagnet position C, then in step 1801, the parameter acquisition calculation unit 222 initializes the indices V, T, and C of the elements of the three-dimensional array to be recalculated to V_S, T_S, and C_S. 18, the parameter acquisition calculation unit 222 recalculates the parameters stored in element I(V,T,C) of the three-dimensional array according to the adjustment amount ΔI of the current amount for each electromagnet. In other words, the parameter acquisition calculation unit 222 recalculates the parameters of the current amount I to be applied to the electromagnet at electromagnet position C at time T when the target energy amount V is set. Note that at this time, the parameter acquisition calculation unit 222 may determine the recalculated current amount I using the current amount I before recalculation and the current adjustment amount ΔI. 18, the parameter acquisition calculation unit 222 judges whether or not the electromagnet position C is the last electromagnet position C_E among the indexes of the elements of the three-dimensional array to be recalculated. If the judgment result of step 1803 is positive, the parameter acquisition calculation unit 222 transfers control to step 1805. If the judgment result of step 1803 is negative, the parameter acquisition calculation unit 222 transfers control to step 1804. 18, the parameter acquisition calculation unit 222 updates the electromagnet position C among the indexes of the elements of the three-dimensional array to be recalculated. For example, if the difference in the electromagnet positions is ΔC, the parameter acquisition calculation unit 222 adds ΔC to the current value of C to obtain a new value of C. Then, the parameter acquisition calculation unit 222 transfers control to step 1802 and recalculates the parameter of the current amount I using the new value of C. 18, the parameter acquisition calculation unit 222 judges whether the time T is the last time T_E of the indexes of the elements of the three-dimensional array to be recalculated. If the judgment result of step 1805 is positive, the parameter acquisition calculation unit 222 transfers control to step 1807. If the judgment result of step 1805 is negative, the parameter acquisition calculation unit 222 transfers control to step 1806. 18, the parameter acquisition calculation unit 222 updates the time T in the index of the element of the three-dimensional array to be recalculated. For example, if the time difference is ΔT, the parameter acquisition calculation unit 222 adds ΔT to the current value of T to obtain a new value of T. The parameter acquisition calculation unit 222 also returns the value of the electromagnet position C to its initial value, C_S. Then, the parameter acquisition calculation unit 222 transfers control to step 1802, and recalculates the parameter of the current amount I using the new value of T and the initialized value of C. 18, the parameter acquisition calculation unit 222 judges whether the target energy amount V is the last target energy amount V_E among the indexes of the elements of the three-dimensional array to be recalculated. If the judgment result of step 1807 is positive, the parameter acquisition calculation unit 222 completes the recalculation for each element of the three-dimensional array. If the judgment result of step 1807 is negative, the parameter acquisition calculation unit 222 transfers control to step 1808. 18, the parameter acquisition calculation unit 222 updates the target energy amount V among the indexes of the elements of the three-dimensional array to be recalculated. For example, if the difference in the target energy amount is ΔV, the parameter acquisition calculation unit 222 adds ΔV to the current value of V to obtain a new value of V. The parameter acquisition calculation unit 222 also returns the value of time T to the initial value T_S, and returns the value of electromagnet position C to the initial value C_S. Then, the parameter acquisition calculation unit 222 transfers control to step 1802, and recalculates the parameter of the current amount I using the new value of V and the initialized values ​​of T and C. 18, the parameter acquisition calculation unit 222 sequentially processes a triple loop using indexes V, T, and C, with the index of the outermost loop being V and the index of the innermost loop being C. Here, the form of the triple loop may be determined appropriately. Furthermore, the parameter acquisition calculation unit 222 may simultaneously perform the triple loop processes in parallel using a technique such as loop unrolling.

[0032] 4.1.4. Display output of estimated cause and adjusted parameter set In step 706 of FIG. 7, after the parameter acquisition calculation unit 222 creates an adjusted parameter set, in step 707, the display output control unit 231 controls to display or output the cause estimated in step 704 and the adjusted parameter set created in step 706. FIG. 19 shows a display screen displayed on some display device (for example, the display or output device 606 shown in FIG. 6) under the control of the display output control unit 231. In step 707, the display screen 1900 may display a chart showing the most recent BP or deviation 301 for each BPM (the chart in the upper left of the display screen 1900 in the example of FIG. 19) and a chart showing the most recent excitation amount B or current amount I for each electromagnet (the chart in the upper right of the display screen 1900 in the example of FIG. 19). These charts may accurately show the information on BP, deviation 301, excitation amount B, and current amount I. Alternatively, these charts may be displayed in some simplified form so that the characteristics of the information on BP, deviation 301, excitation amount B, and current amount I are visually easy to understand for an operator, etc. In step 707, the display screen 1900 may have a pre-adjustment parameter set file output icon 1901. When the pre-adjustment parameter set file output icon 1901 is clicked using a mouse or the like, a file (e.g., a CSV file) indicating the pre-adjustment parameter set (e.g., a parameter set in a three-dimensional array as described with reference to FIG. 18) is output. The operator or the like can separately view the file using appropriate software. In step 707, the display screen 1900 may have an estimated cause display field 1903. The estimated cause display field 1903 may display the cause estimated in step 704. The estimated cause display field 1903 may display whether a value is set or reset for each component (item) of the estimated cause vector 501 shown in FIG. 5. Here, the display may show the value of each component (item) itself, or may show the estimated cause in words. Instead of the estimated cause display field 1903, or together with the estimated cause display field 1903, an estimated cause output icon may be present. When the estimated cause output icon is clicked with a mouse or the like, information held by the estimated cause vector may be output in the form of a file. In step 707, the display screen 1900 may display a chart (in the example of FIG. 19, the chart in the lower right of the display screen 1900) showing the parameters of the excitation amount B or the current amount I for each electromagnet created in step 706. The display screen 1900 may also display a chart (in the example of FIG. 19, the chart in the lower left of the display screen 1900) showing the BP or deviation 301 for each BPM, which is assumed when the parameters of the excitation amount B or the current amount I for each electromagnet created in step 706 are applied. These charts may accurately show the information of the BP, the deviation 301, the excitation amount B, and the current amount I. Alternatively, these charts may be displayed in some simplified form so that the characteristics of the information of the BP, the deviation 301, the excitation amount B, and the current amount I are visually easy to understand for the operator, etc. In step 707, the display screen 1900 may have an adjusted parameter set file output icon 1902. When the adjusted parameter set file output icon 1902 is clicked using a mouse or the like, a file (e.g., a CSV file) indicating the adjusted parameter set (e.g., the parameter set in the three-dimensional array described in FIG. 18) is output. The operator or the like can separately view the file using appropriate software. The operator or the like can also further modify the parameters included in the file. This step 707 may be a parameter display output step. In step 708 of FIG. 7, the display output control unit 231 performs control so as to display or output the cause estimated in step 704. In step 708 of FIG. 7, under the control of the display output control unit 231, some or all of the following may be displayed on the display screen: a graph showing the most recent BP or deviation 301 for each BPM (in the example of FIG. 19, the graph in the upper left corner of the display screen 1900), a graph showing the most recent excitation amount B or current amount I for each electromagnet (in the example of FIG. 19, the graph in the upper right corner of the display screen 1900), a pre-adjustment parameter set file output icon 1901, and an estimated cause display column 1903. 7, the deviation of the beam trajectory 173 from the design beam trajectory is often complex. Therefore, in this case, the information processing system 101 may only display the cause estimated in step 704 and data indicating the current state to the operator, etc. After this, it is assumed that a person with expert knowledge of the accelerator 170 manually adjusts the file (e.g., a CSV file) indicating the pre-adjustment parameter set. If the display output control unit 231 controls to display or output the cause estimated in step 704 and the adjusted parameter (set) obtained in step 706 using a display screen 1900 as shown in Fig. 19, the estimated cause and the adjusted parameter (set) can be communicated to the operator of the accelerator 170 in an easily understandable manner. Furthermore, the display output control unit 231 can provide the parameter sets before and after adjustment in the form of a file if necessary to the operator of the accelerator 170. Therefore, the operator of the accelerator 170 can set the parameter set obtained in the form of a file as it is in the accelerator control device 240, or can further modify the parameter set obtained in the form of a file and then set the modified parameter set in the accelerator control device 240.

[0033] 4.1.5. Decision on warm-up operation (C) In step 709 of FIG. 7, the warm-up operation determination unit 223 determines whether or not to perform warm-up operation in the accelerator 170, and if warm-up operation is to be performed, determines what mode of warm-up operation to perform, based on the information on the BP, electromagnet, sensor, or system acquired in steps 701 and 703. In addition, a method similar to the determination regarding the warm-up operation described below can be used to determine whether or not to perform cooling operation in the accelerator 170, and to determine the mode of cooling operation if cooling operation is performed, which will be shown later as a modified example. Three methods will be described below as examples of the warm-up operation determination process executed by the warm-up operation determination unit 223. The three methods are a method based on cause estimation, a method based on DB search, and a method using a warm-up operation determination model. The warm-up operation determination unit 223 may execute any one of these three methods. Alternatively, the warm-up operation determination unit 223 may selectively execute any one of these three methods depending on whether a predetermined condition is satisfied. Alternatively, the warm-up operation determination unit 223 may ensemble (for example, by majority vote or average) the determination results of the presence or absence of warm-up operation and the determination results of the mode of warm-up operation obtained by each of the three methods to obtain a final determination result regarding the warm-up operation. Furthermore, the warm-up operation determination unit 223 may make a determination regarding the warm-up operation using a method other than these three methods. This step 709 may be a warm-up operation determination step. In addition, in the warm-up operation determination process described below, the warm-up operation determination unit 223 determines the remaining time of the warm-up operation (the duration from the current time) as the mode of the warm-up operation when the warm-up operation is performed. Alternatively, the warm-up operation determination unit 223 may determine the time period during which the warm-up operation is performed as the mode of the warm-up operation when the warm-up operation is performed. In this case, in the record held by the database (DB) 260 (or each piece of information before being stored in the DB 260), among the components (items) shown in Fig. 4, the component (item)

[14] may have information on the time period during which the warm-up operation is performed. In this modified example in which the time period for performing the warm-up operation is determined, it is expected that the warm-up operation will be performed appropriately, for example, during a time period when the air temperature is lower.

[0034] 4.1.5.1. Decision on warm-up operation based on cause estimation (C-1) FIG. 20 shows a flowchart of the warm-up operation determination based on cause estimation that can be performed by the warm-up operation determination unit 223 in step 709 in FIG. In step 2001 of FIG. 20, the warm-up operation determination unit 223 estimates the cause of deviation of BP information from the adjustment-free range 302 in any of the BPMs 181 to 184. The mode of the cause estimation process may be the same as that already described for the cause estimation unit 221. Alternatively, in step 2001 of FIG. 20, the warm-up operation determination unit 223 may request the cause estimation process to the cause estimation unit 221 and receive the estimation result from the cause estimation unit 221. Alternatively, in step 2001 of FIG. 20, the warm-up operation determination unit 223 may use the result of the cause estimation process performed by the cause estimation unit 221 in step 704. In step 2002 in FIG. 20, the warm-up operation determination unit 223 determines whether or not the causes estimated in the estimation result obtained in step 2001 are caused by the cooling water temperature of the accelerator 170. For example, in each component (item) included in the estimated cause vector in the format shown in FIG. <4> and <5> These two components (items) may be derived from the cooling water temperature of the accelerator 170. More specifically, in step 2002 of FIG. 20, the warm-up operation determination unit 223 may make the determination result of step 2002 positive if it is determined that the cooling water temperature is lower than a predetermined temperature threshold or that the elapsed time after the start-up or restart of the particle beam therapy system 160 is shorter than a predetermined time threshold. If the determination result of step 2002 is positive, the warm-up operation determination unit 223 shifts control to step 2003. If the determination result of step 2002 is negative, the warm-up operation determination unit 223 shifts control to step 2004. In step 2003 in FIG. 20, the warm-up operation determination unit 223 determines that a warm-up operation is being performed as a result of the warm-up operation determination process, and appropriately determines the remaining time of the warm-up operation (the duration from the current time). In step 2004 in FIG. 20, the warm-up operation determination unit 223 determines that the warm-up operation is not performed as a result of the warm-up operation determination process. By using the warm-up operation judgment process based on cause estimation as shown in Figure 20, if it is determined that an event in which a warm-up operation is an effective countermeasure, such as low cooling water level used in the accelerator 170, is included in the causes of the beam trajectory 173 (BP and deviation 301 from normal values) deviating from the adjustment-free range 302 from the designed beam trajectory, the information processing system 101 can recommend to an operator, etc., that a warm-up operation be performed, or set the accelerator control device 240 to automatically execute a warm-up operation.

[0035] 4.1.5.2. Determining whether to warm up the engine by searching the database (C-2) FIG. 21 shows a flowchart of the warm-up operation determination by the warm-up operation determination unit 223 through DB search in step 709 in FIG. In step 2101 of Fig. 21, the warm-up operation determination unit 223 compares the component (item) information in each record 401 in the format shown in Fig. 4 from the database (DB) 260 with the component (item) information obtained from the information on the BP, electromagnet, sensor, and system acquired in step 701 or step 703. The components (items) compared here may be, for example, information corresponding to some or all of the nine components (items) consisting of [1] to [9] among the components (items) shown in Fig. 4. Furthermore, the warm-up operation determination unit 223 may perform a comparison process on components (items) other than the components (items) listed above. 21, the warm-up operation determination unit 223 selects a record having many similarities with each piece of information on the BP, electromagnet, sensor, and system (information linked to the deviation of the beam trajectory 173 (BP, deviation 301) from the adjustment-free range 302) acquired in step 701 or step 703 from among the records 401 compared in step 2101. For example, the warm-up operation determination unit 223 may obtain a score representing the similarity for each record by scoring the difference obtained in the comparison process in step 2101 for each of the nine components (items) described above, and select the record having the highest score in step 2102. In step 2103 of FIG. 21, the warm-up operation judgment unit 223 determines, based on the two components (items)

[13] and

[14] contained in the record selected in step 2102, whether or not warm-up operation is being performed and the nature of the warm-up operation (e.g., the remaining time of the warm-up operation (the duration from the current time)) as a result of the warm-up operation judgment process. The above-described process of determining whether or not to warm up the accelerator by searching the database makes it relatively easy to determine whether or not to warm up the accelerator by utilizing the accelerator's past operation history information.

[0036] 4.1.5.3. Warm-up operation judgment using warm-up operation judgment model (C-3) Fig. 22 shows inputs and outputs of the warm-up operation judgment model 2200 when the warm-up operation judgment unit 223 executes the warm-up operation judgment model 2200 when performing the warm-up operation judgment process in step 709 in Fig. 7. The warm-up operation judgment unit 223 realizes the warm-up operation judgment model 2200 based on a program and a model parameter group for the warm-up operation judgment model 2200. Note that when realizing the warm-up operation judgment model 2200, a part or the whole may be realized by hardware rather than by program control. The warm-up operation judgment model 2200 receives as input the BP, electromagnet, sensor, and system information (information linked to the deviation of the beam trajectory 173 (BP, deviation 301) from the adjustment-free range 302) acquired in step 701 or step 703, and outputs the presence or absence of warm-up operation and the state of warm-up operation (for example, the remaining time of warm-up operation (continuation time from the current time)) as a result of the warm-up operation judgment process. In FIG. 22, the inputs to the warm-up operation judgment model 2200 may include BP information (information corresponding to [9] in FIG. 4), electromagnet information (information corresponding to [8] in FIG. 4), sensor information (information corresponding to [1], [2], [5], and [7] in FIG. 4), and system information (information corresponding to [3], [4], and [6] in FIG. 4). The output of the parameter acquisition model 1500 may include whether or not warm-up operation is performed (information corresponding to

[13] in FIG. 4) and the state of the warm-up operation (e.g., the remaining time of the warm-up operation (the duration from the current point in time)) (information corresponding to

[14] in FIG. 4). 23 shows a flowchart of machine learning for training model parameters to construct the warm-up operation judgment model 2200. This machine learning may be executed by the information processing system 101 or may be executed by some system other than the information processing system 101. In step 2301 in FIG. 23, the system that performs machine learning reads one record from a record group to be used for machine learning, among the records held by the DB 260 (records in the format shown in FIG. 4). In step 2302 of FIG. 23, the system performing machine learning inputs some or all of the nine components (items) consisting of [1] to [9] of the components (items) contained in the record read in step 2301 into the warm-up operation judgment model 2200. In step 2303 of Fig. 23, the system that executes machine learning obtains an output from the warm-up operation judgment model 2200 according to the input of each component (item) to the warm-up operation judgment model 2200 performed in step 2302. Here, the output from the warm-up operation judgment model 2200 is in the form of the presence or absence of warm-up operation (information corresponding to

[13] in Fig. 4) and the state of the warm-up operation (for example, the remaining time of the warm-up operation (the duration from the current time)) (information corresponding to

[14] in Fig. 4) as a result of the warm-up operation judgment process. In step 2304 of FIG. 23, the system performing machine learning uses information on two components (items),

[13] and

[14] , of the components (items) contained in the record read in step 2301, as training data. 23, the system that executes machine learning treats as an error the difference between the presence or absence of warm-up operation and the state of the warm-up operation (e.g., the remaining time of the warm-up operation) obtained in step 2303 and the presence or absence of warm-up operation and the state of the warm-up operation (e.g., the remaining time of the warm-up operation) that is the teacher data obtained in step 2304. Then, the system that executes machine learning adjusts (trains) the model parameters of the warm-up operation judgment model 2200 by an error backpropagation method or the like using the error. In step 2306 of Fig. 23, the system that performs machine learning judges whether all records included in the record group to be used for machine learning have been read out from among the records held by DB 260. If the result of the judgment indicates that there are records that have not yet been read out in step 2301 to be used for machine learning, the system that performs machine learning returns control to step 2301, determines the next record to be read out from among the records that have not yet been read out in step 2301 to be used for machine learning, and reads out that record. If the result of the judgment indicates that all records to be used for machine learning have been read out in step 2301, the machine learning process of the warm-up operation judgment model 2200 is completed. FIG. 24 shows a flowchart of the warm-up operation determination process in the case where the warm-up operation determination unit 223 executes the warm-up operation determination model 2200 when performing the warm-up operation determination process in step 709 in FIG. In step 2401 of FIG. 24, the warm-up operation judgment unit 223 inputs information corresponding to some or all of the nine components (items) consisting of [1] to [9] from the information on BP, electromagnet, sensor, and system acquired in step 701 or step 703 (information linked to the deviation of the beam trajectory 173 (BP, deviation 301) from the adjustment-free range 302) into the warm-up operation judgment model 2200. In step 2402 in FIG. 24, the warm-up operation judgment unit 223 obtains, as output from the warm-up operation judgment model 2200, the presence or absence of warm-up operation (information corresponding to

[13] in FIG. 4) and the state of warm-up operation (e.g., the remaining time of warm-up operation) (information corresponding to

[14] in FIG. 4). The warm-up operation judgment process using a model trained by machine learning as described above can realize warm-up operation judgment that fully reflects the accelerator's past operating history information.

[0037] 4.1.5.4. Display output of warm-up operation judgment result 7, the warm-up operation determination unit 223 determines whether or not it has been determined that a warm-up operation should be performed (warm-up operation is performed) as a result of the warm-up operation determination process in step 709. If the determination result in step 710 is positive, the warm-up operation determination unit 223 shifts control to step 711. In step 711 of FIG. 7, the display output control unit 231 controls to display or output recommended information (recommendation information) regarding the warm-up operation, which is the result of the warm-up operation determination process in step 709, including a message indicating that the warm-up operation should be performed (the warm-up operation is performed) and the state of the warm-up operation (for example, the remaining time of the warm-up operation). Here, when the display output control unit 231 controls to display or output the recommended information (recommendation information) regarding the warm-up operation in step 711, the display output control unit 231 may control to display the recommended information (recommendation information) regarding the warm-up operation on the display screen 1900 of FIG. 19. Alternatively, when the display output control unit 231 controls to display or output the recommended information (recommendation information) regarding the warm-up operation in step 711, the display output control unit 231 may control to display or output the recommended information (recommendation information) regarding the warm-up operation on a display screen other than the display screen 1900 of FIG. If the display output control unit 231 controls to display or output the above-mentioned recommended information (recommended information) regarding the warm-up operation, the recommended contents of the warm-up operation for the accelerator 170 can be communicated to the operator of the accelerator 170 in an easy-to-understand manner. Furthermore, the display output control unit 231 may provide the recommended information (recommended information) regarding the warm-up operation in the form of a file if necessary to an operator, etc. of the accelerator 170. In this way, the operator, etc. of the accelerator 170 can set the recommended contents (recommended information) regarding the warm-up operation obtained in the form of a file as is in the accelerator control device 240, or can further modify the recommended contents (recommended information) regarding the warm-up operation obtained in the form of a file and then set the modified recommended contents (recommended information) regarding the warm-up operation in the accelerator control device 240.

[0038] 4.1.6.Setting up the accelerator control device 7, in step 712 after step 707 or step 711, the parameter setting unit 232 judges whether or not an instruction to set the contents displayed or output in step 707 or step 711 (parameter set for the electromagnets or warm-up operation parameters), or contents further modified by the operator, in the accelerator control device 240, has been received from the operator, etc. If the judgment result in step 712 is positive, the parameter setting unit 232 shifts control to step 713. 7, the parameter setting unit 232 sets the contents instructed by the operator or the like as the object to be set (parameter set for the electromagnet, or warm-up operation parameters) in the accelerator control device 240. For example, if a parameter set for the electromagnet is to be set, the parameter setting unit 232 stores the parameter set for the electromagnet to be set in the parameter storage device 241. For example, if a parameter for the warm-up operation (whether or not to perform the warm-up operation and the mode of the warm-up operation (for example, the remaining time of the warm-up operation (the duration from the current time)) is to be set), the parameter setting unit 232 sets the warm-up operation parameters to be set in the accelerator control device 240, and causes the accelerator 170 to execute the warm-up operation according to the set contents. Step 713, which is performed after the determination result of step 712 is positive, may be a parameter post-instruction setting step. In this way, by allowing the operator or the like to confirm and examine the displayed or output contents once and further modify the parameters as necessary, the operator or the like who has specialized knowledge about the accelerator 170 can be involved in the setting process for the accelerator control device 240, and therefore the control of the accelerator 170 becomes more appropriate. Incidentally, in response to a selection by an operator or the like of a target amount of energy V to be given to the beam 171, the accelerator controller 240 selects a group of elements corresponding to the selected target amount of energy V from the parameter set. Then, as time T progresses, the accelerator controller 240 sequentially applies parameters of the elements of I[V,T,C] (for example, a parameter of the amount of current I) to the electromagnet at the electromagnet position C. In this way, the parameter set is stored (under the control of the parameter setting unit 232 in the information processing system 101) in the parameter storage device 241 which can be accessed from the accelerator control device 240 at a relatively high speed, and the accelerator control device 240 then applies the parameters stored in the parameter storage device 241 to the accelerator 170 as appropriate. Therefore, even when the accelerator control device 240 changes settings such as changing the target energy amount V, or when the time T is initialized to initialize the acceleration process of the beam 171, the accelerator control device 240 can quickly apply the parameters to the accelerator 170.

[0039] 4.2. Processing for determining when to adjust parameters Among the processes that can be realized by the embodiments of the present disclosure, a process (adjustment time determination process) for determining the time (parameter adjustment time) to adjust (create) parameters for controlling the accelerator 170 in order to bring the beam orbit 173 closer to the design beam orbit (bring the BP or deviation 301 closer to the normal value) in a future period is described in Figures 25 to 31. It is assumed that the following adjustment time determination process is also applied to the accelerator 170 similar to the accelerator 170 to which the cause estimation process, parameter adjustment (creation) process, and warm-up operation determination process already described are applied. In the case of an accelerator having a plurality of BPMs 181 to 184 like the accelerator 170, the adjustment time determination process can be performed taking into account the distortion of the beam orbit 173. However, the following adjustment time judgment process is sufficient if it can handle the time series information of BP information (hereinafter, sometimes referred to as "BP time series information"). Therefore, since it is not necessarily required that an accelerator has multiple BPMs, the adjustment time judgment process can also be applied to an accelerator that has only one BPM. In this way, the adjustment time determination process is applicable to a wide range of accelerators. In the adjustment time determination process described below, similarly to the cause estimation process, parameter adjustment (creation) process, and warm-up operation determination process described above, information corresponding to some or all of the components (items) [1] to [9] among the components (items) shown in FIG. 4 is used as information to be input in the process. In addition, in the adjustment time determination process described below, past information on the component (item)

[11] "whether or not a parameter is adjusted" may also be used as information to be input in the process. By using such information indicating past parameter adjustment times in the adjustment time determination process, the adjustment time determination process can be realized such that, for example, "if the previous parameter adjustment time was a day belonging to midsummer and the current date and outside temperature are those of autumn, it is determined that the time to perform the next parameter adjustment is the time when autumn changes to winter." Fig. 25 shows a main flowchart of the adjustment timing determination process. Note that Figs. 27 to 31 are drawings that further explain the estimation process of time-series information of BP information for a future period, which is performed in step 2501 of Fig. 25.

[0040] 4.2.1. Estimation of BP time series information for future periods (BP time series information) (D) In step 2501 in FIG. 25, the adjustment timing determination unit 224 estimates time-series information of BP information (BP time-series information) for each of the BPMs 181 to 184 in a future period (hereinafter, sometimes referred to as an estimated period or a subsequent period). Two methods will be described below as examples of the BP time-series information estimation process executed by the adjustment timing determination unit 224. The two methods are a method using a regression analysis model and a method using a BP time-series estimation model. The adjustment time determination unit 224 may execute one of these two methods. Alternatively, the adjustment time determination unit 224 may selectively execute one of these two methods depending on whether a predetermined condition is satisfied. Alternatively, the adjustment time determination unit 224 may ensemble (for example, take the average) the estimation results of the BP time series information obtained by each of these two methods to obtain a final estimation result of the BP time series information. FIG. 26 illustrates the technical background of how the adjustment timing determination unit 224 estimates BP time-series information in step 2501 when determining the parameter adjustment timing for a future period (estimated period, subsequent period). 26, the adjustment time determination unit 224 estimates BP time series information 2601 for a future period (estimated period, subsequent period), whereby the adjustment time determination unit 224 can determine the time when the BP information (or deviation 301) for each time T included in the BP time series information 2601 will deviate from the adjustment-free range 302. The adjustment time determination unit 224 can set the determined time as the parameter adjustment time 2603.

[0041] 4.2.1.1. Estimation of BP time series information using regression analysis model (D-1) Fig. 27 is a flowchart showing the generation process of a regression analysis model when the adjustment timing determination unit 224 performs the process of estimating BP time-series information using a regression analysis model in step 2501 in Fig. 25. The generation process of the regression analysis model shown in Fig. 27 may be executed by the information processing system 101 or a system different from the information processing system 101. In step 2701 of FIG. 27, the system that performs the process of generating a regression analysis model obtains time series information for each piece of information, including BP, electromagnetic, sensor, and system, from each record 401 in the database (DB) 260 (or from the data accumulation server 250 or the sensor data collection device 280, if the most recent information that has not yet been accumulated in the DB 260 is also to be used). This step 2701 may be a step of acquiring beam position time-series information. Also, step 2701 may be a step of acquiring condition time-series information. In step 2702 of FIG. 27, the system performing the regression analysis model generation process generates a regression analysis model using the time-series information of the BP, electromagnetic, sensor, and system information acquired in step 2701. Here, the system performing the regression analysis model generation process sets information corresponding to some or all of the components (items) [1] to [8] among the components (items) included in the record 401 in the format shown in FIG. 4 as explanatory variables in the regression analysis model. The system performing the regression analysis model generation process may also set the elapsed time from the already performed parameter adjustment, which is derived from the component (item)

[11] among the components (items) included in the record 401 in the format shown in FIG. 4, as an explanatory variable in the regression analysis model. The system performing the regression analysis model generation process sets information corresponding to the component (item) [9] among the components (items) included in the record 401 in the format shown in FIG. 4 as an objective variable in the regression analysis model. FIG. 28 shows a flowchart of the estimation process when the adjustment timing determination unit 224 performs the estimation process of the BP time-series information using a regression analysis model in step 2501 of FIG. 28, the adjustment time determination unit 224 may obtain information corresponding to the component (item)

[11] among the components (items) in the format shown in FIG. 4 from each record 401 in the DB 260 (from the data accumulation server 250 or the sensor data collection device 280, etc., if the most recent information not yet accumulated in the DB 260 is also used). Then, the adjustment time determination unit 224 may determine information on the time (date or time) when the parameters for controlling the accelerator 170 (for example, the parameter of the current amount I for each of the electromagnets 191-194) were last adjusted, based on the information corresponding to the component (item)

[11] obtained above. The information on the time (date or time) when the last adjustment was made may be used as an explanatory variable of the regression analysis model. This step 2801 may be an adjustment time information acquisition step. In steps 2802 to 2809 in Fig. 28, the adjustment time determination unit 224 may provisionally determine information corresponding to some or all of the components (items) [1] to [8] in the period (future period, estimated period, subsequent period) after the current time 2602 in Fig. 26, and then use the provisionally determined variables as explanatory variables of the regression analysis model. In this way, the adjustment time determination unit 224 obtains information corresponding to the component (item) [9] in the period (future period, estimated period, subsequent period) after the current time 2602 in Fig. 26 as the response variable of the regression analysis model, and sets it as the estimation result of the BP time series information. In the example of Fig. 28, the adjustment time determination unit 224 stores, as the estimation result of the BP time series information, the element ΔX(T,M) of the two-dimensional array regarding the deviation 301 between the BP information and the normal value, when the time included in the period after the current time 2602 is T and the BPM position is M. The adjustment timing determination unit 224 may, as necessary, change the contents of the provisional determination of information corresponding to each component (item) of [1] to [8] above for the period after the present 2602 in Fig. 26 (future period, estimated period, subsequent period), and re-estimate the BP time-series information. Each step will be explained below. 28, the adjustment timing determination unit 224 initializes the time T of the index (T, M) of the two-dimensional array of the deviation 301 to be estimated. For example, the present 2602 or the time immediately following the present 2602 is set as T_N, and the time T, which is the index, is set as T_N. 28, the adjustment timing determination unit 224 obtains the elapsed time from the time of the last adjustment to time T, which is an index of the two-dimensional array, based on information about the time (date or time) when the parameters for controlling the accelerator 170 (for example, the parameters of the current amount I for each of the electromagnets 191-194) obtained in step 2801 were last adjusted. This elapsed time may be one of the explanatory variables of the regression analysis model. Note that it is not essential to use this elapsed time as an explanatory variable of the regression analysis model, so the estimation process of the BP time series information by the regression analysis model may be performed without using this elapsed time. In step 2804 of FIG. 28, the adjustment time determination unit 224 may provisionally determine information corresponding to some or all of the components (items) [1] to [8] above for time T, which is an index of the two-dimensional array, and use the provisionally determined information as an explanatory variable of the regression analysis model. In step 2805 of FIG. 28, the adjustment timing determination unit 224 obtains, as the output (objective variable) from the regression analysis model, BP information (pre-adjustment BP) measured for each BPM, which is information corresponding to the components of [9] above, or information on the deviation of BP from the normal value. This step 2805 may be a beam position time series information estimation step. In step 2806 in FIG. 28, the adjustment timing decision section 224 sets (stores) a two-dimensional array ΔX(T,M) with M as the BPM position based on the BP information or deviation at time T obtained in step 2805. 28, the adjustment time determination unit 224 determines whether the time T, which is an index of the two-dimensional array, has reached the end of the period (future period, estimated period, subsequent period) for which the BP time series information is estimated. If the determination result of step 2807 is positive, the adjustment time determination unit 224 transfers control to step 2809. If the determination result of step 2807 is negative, the adjustment time determination unit 224 transfers control to step 2808. 28, the adjustment time determination unit 224 updates the time T, which is an index of the two-dimensional array. For example, when the time difference is ΔT, the adjustment time determination unit 224 adds the difference ΔT to the current time T to obtain the new time T. After step 2808, the adjustment time determination unit 224 returns control to step 2803 and continues calculating the element ΔX(T,M) of the two-dimensional array for the new time T. 28, the adjustment time determination unit 224 outputs a two-dimensional array ΔX(T,M) over the period (future period, estimation period, subsequent period) for estimating the BP time-series information. After step 2809, the adjustment time determination unit 224 transfers control to step 2810. In step 2810 in Fig. 28, the adjustment time judgment unit 224 judges whether to redo the estimation process of the BP time series information using new provisional determinations as the contents of the provisional determinations of information corresponding to some or all of each of the components (items) [1] to [8] above over the period to be estimated of the BP time series information (future period, estimated period, subsequent period). If the judgment result in step 2810 is positive, the adjustment time judgment unit 224 returns control to step 2802 and continues the calculation of the element ΔX(T,M) of the two-dimensional array using the new provisional determinations. If the judgment result in step 2810 is negative, the estimation process of the BP time series information using the regression analysis model shown in Fig. 28 is completed. In addition, the provisional determination of information corresponding to some or all of the components (items) [1] to [8] above, which is performed in step 2804 of Fig. 28, may be performed using various methods. For example, the adjustment time determination unit 224 may perform the provisional determination above for the period (future period, estimation period, subsequent period) for which the BP time-series information is estimated, by using an average value of information for the same period over the past few years. Alternatively, the adjustment time determination unit 224 may set a provisional determination of information corresponding to each of the components (items) [1] to [8] above, which indicates a marginal situation in the accelerator 170, for the period (future period, estimation period, subsequent period) for which the BP time-series information is estimated. 28 shows that the adjustment time determination unit 224 performs sequential processing for time T. However, the adjustment time determination unit 224 may perform processing using a regression analysis model simultaneously for time T. The above-described estimation process of BP time series information using a regression analysis model has the advantage that it is relatively easy to verify the relationship between the explanatory variables and the target variable, allowing operators and others to obtain estimation results of BP time series information while grasping the correlation between variables.

[0042] 4.2.1.2. Estimation of BP time series information using BP time series estimation model (D-2) Fig. 29 shows input and output of the BP time series estimation model 2900 when the adjustment time determination unit 224 executes the BP time series estimation model 2900 when performing the adjustment time determination process in step 2501 in Fig. 25. The adjustment time determination unit 224 realizes the BP time series estimation model 2900 based on a program and a group of model parameters for the BP time series estimation model 2900. Note that when realizing the BP time series estimation model 2900, a part or the whole may be realized by hardware rather than by program control. The BP time series estimation model 2900 receives, as inputs to the model, information on the parameter adjustment timing in the period preceding the present 2602 (hereinafter, this period may be referred to as the preceding period. In FIG. 29, it is abbreviated as [previous]) (information corresponding to the component (item)

[11] in FIG. 4), BP information for the preceding period (information corresponding to [9] in FIG. 4), electromagnet information for the preceding period (parameters for controlling the electromagnets for each of the electromagnets 191 to 194 (for example, parameters for the amount of current I and the amount of excitation B)) (information corresponding to [8] in FIG. 4), and sensor information for the preceding period (information corresponding to

[11] in FIG. 4). 29), system information of the preceding period (information corresponding to [3], [4], and [6] in FIG. 4), (provisionally determined) electromagnet information (information corresponding to [8] in FIG. 4) of the period following the present 2602 (period for estimating BP time series information. Hereinafter, this may be referred to as the following period. In FIG. 29, it is abbreviated as [following].), (provisionally determined) sensor information of the following period (information corresponding to [1], [2], [5], and [7] in FIG. 4), and (provisionally determined) system information of the following period (information corresponding to [3], [4], and [6] in FIG. 4). In addition, the BP time series estimation model 2900 has BP information (BP time series information) of the following period (information corresponding to [9] in FIG. 4) as an output from the model. FIG. 30 shows a flowchart of machine learning for training model parameters to construct the BP time series estimation model 2900. This machine learning may be executed by the information processing system 101, or may be executed by some system other than the information processing system 101. In the flowchart of machine learning for the BP time series estimation model 2900 shown in FIG. 30, the system that executes machine learning selects two consecutive periods in the past, treats the two periods as a pair of a preceding period and a following period, and selects a record group corresponding to the preceding period and a record group corresponding to the following period as a record group to be used for machine learning. Then, the system that executes machine learning repeats the selection of two consecutive periods in the past a predetermined number of times to repeat machine learning for the BP time series estimation model 2900. Each step will be described below. In step 3001 of Figure 30, the system performing machine learning reads out a set of records corresponding to a preceding period and a set of records corresponding to a subsequent period from among the records held by DB260 (records in the format shown in Figure 4) as sets of records to be used for machine learning. In step 3002 of FIG. 30, the system performing machine learning inputs some or all of the 10 components (items) consisting of [1] to [9] and

[11] from the set of records corresponding to the preceding period read in step 3001 into the BP time series estimation model 2900. In addition, in step 3002 of FIG. 30, the system performing machine learning inputs some or all of the eight components (items) consisting of [1] to [8] among the components (items) of the record group corresponding to the subsequent period read in step 3001 to the BP time series estimation model 2900. 30, the system performing machine learning obtains an output from the BP time series estimation model 2900 in response to the input of each component (item) to the BP time series estimation model 2900 performed in step 3002. Here, the output from the BP time series estimation model 2900 is in the form of BP information (pre-adjusted BP) measured for each BPM in the subsequent period, or BP deviation from the normal value (information corresponding to [9] in FIG. 4), as a result of the estimation process of the BP time series information. In step 3004 of FIG. 30, the system performing machine learning sets the information of component (item) [9] among the components (items) contained in the record group corresponding to the subsequent period read in step 3001 as training data. 30, the system performing machine learning treats as an error the difference between the BP information (pre-adjustment BP) measured for each BPM in the subsequent period or the information on the deviation of BP from the normal value obtained in step 3003, and the BP information (pre-adjustment BP) measured for each BPM or the information on the deviation of BP from the normal value, which is the teacher data obtained in step 3004. Then, the system performing machine learning adjusts (trains) the model parameters of the BP time series estimation model 2900 by an error backpropagation method or the like using the error. In step 3006 of Fig. 30, the system that executes machine learning judges whether or not the operation of selecting two consecutive periods in the past for records held by the DB 260, treating the two periods as a pair of a preceding period and a following period, and selecting a record group corresponding to the preceding period and a record group corresponding to the following period as a record group to be used for machine learning has been executed for all planned pairs. If the judgment result in step 3006 is negative, the system that executes machine learning transfers control to step 3001 and executes machine learning of the BP time series estimation model 2900 for a new pair of a preceding period and a following period. If the judgment result in step 3006 is positive, the machine learning process of the BP time series estimation model 2900 is completed. FIG. 31 shows a flowchart of the adjustment time judgment process in the case where the adjustment time judgment unit 224 executes the adjustment time judgment process in step 2501 of FIG. 31, the adjustment time determination unit 224 acquires information (time-series information) corresponding to some or all of the 10 components (items) consisting of [1] to [9] and

[11] from the information on the BP, electromagnet, sensor, and system in the period preceding the present 2602 (hereinafter, sometimes referred to as the preceding period). Note that the adjustment time determination unit 224 may acquire this information from the DB 260, the data accumulation server 250, the sensor data collection device 280, or the like. This step 3101 may be a step of acquiring beam position time series information. Also, step 3101 may be a step of acquiring condition time series information. Furthermore, step 3101 may be a step of acquiring adjustment timing information. In step 3102 of FIG. 31, the adjustment timing determination unit 224 inputs the information (time series information) corresponding to some or all of the above components (items) [1] to [9] and

[11] of the preceding period, which is the information acquired in step 3101, to the BP time series estimation model 2900. In step 3103 of FIG. 31, the adjustment timing determination unit 224 provisionally determines some or all of the information (time series information) corresponding to the components (items) [1] to [8] above for each time T in the period following the present 2602 (hereinafter sometimes referred to as the subsequent period), and inputs the provisionally determined information (time series information) to the BP time series estimation model 2900. 31, the adjustment time determination unit 224 obtains information (time series information) of the components (items) in [9] above for each time T in the subsequent period as the output of the BP time series estimation model 2900. That is, the adjustment time determination unit 224 obtains, as the output of the BP time series estimation model 2900, information (time series information) corresponding to BP information measured for each BPM (pre-adjustment BP) or information on the deviation of BP from the normal value (time series information). This step 3104 may be a beam position time series information estimation step. 31, the adjustment time determination unit 224 sets each of the elements ΔX(T,M) of a two-dimensional array that stores an estimated value of the BP time series information (for example, an estimated value of the deviation 301) with T as the time and M as the BPM position based on the information (time series information) of the components (items) of the above [9] for the subsequent period obtained in step 3104. Then, the adjustment time determination unit 224 outputs each of the elements ΔX(T,M) of the two-dimensional array as a result of the estimation process of the BP time series information. In step 3106 of Fig. 31, the adjustment time judgment unit 224 judges whether or not to perform the estimation process of the BP time series information with a new provisional determination as the content of the provisional determination made in step 3103 (the above contents of [1] to [8] for the subsequent period). If the judgment result in step 3106 is positive, the adjustment time judgment unit 224 transfers control to step 3103 and redoes the estimation process of the BP time series information with the new provisional determination. If the judgment result in step 3106 is negative, the estimation process of the BP time series information using the BP time series estimation model 2900 is completed. The provisional determination of information corresponding to each of the components (items) [1] to [8] above, which is performed in step 3103 of Fig. 31, may be performed by various methods. For example, the adjustment time determination unit 224 may perform the provisional determination above for the period (subsequent period) for which the BP time-series information is estimated, based on the average value of information for the same period over the past few years. Alternatively, the adjustment time determination unit 224 may set provisional determination of information corresponding to each of the components (items) [1] to [8] above, which indicates a marginal situation in the accelerator 170, for the period (subsequent period) for which the BP time-series information is estimated. The estimation process for BP time series information using a model trained by machine learning as described above can realize estimation that fully reflects the accelerator's past operating history information.

[0043] 4.2.2. Determining the timing of parameter adjustment using estimated BP time series information In step 2501 of FIG. 25, the adjustment time determination unit 224 performs an estimation process of BP time series information for the period after the present 2602 (future period, estimated period, subsequent period), and then in step 2502, the adjustment time determination unit 224 determines the time (parameter adjustment time) to adjust parameters for controlling the accelerator 170 (for example, parameters of the current amount I and excitation amount B for each of the electromagnets 191-194) for the period after the present 2602 (future period, estimated period, subsequent period). The adjustment timing determination section 224 may determine, as the parameter adjustment timing 2603, the time T at which the BP information or deviation 301 of any of the BPMs 181 to 184 in the estimated BP time-series information 2601 deviates from the adjustment-free range 302. 26, the estimated BP time series information 2601 may be expected to include seasonal fluctuations, time zone fluctuations, etc. Therefore, the adjustment timing determination unit 224 may calculate time series information (trend value) indicating the trend of the estimated BP time series information 2601, and set the time T at which the trend value deviates from the adjustment-free range 302 as the parameter adjustment timing 2603. Methods for calculating the trend value include a method using a regression line or regression curve, and a method using a moving average line. This step 2502 may be an adjustment timing determination step.

[0044] 4.2.3. Display and output of estimated BP time series information and parameter adjustment timing In step 2503 of FIG. 25, the display output control unit 231 may perform control so as to display or output the BP time-series information obtained in step 2501 or the parameter adjustment timing obtained in step 2502. For example, in step 2503 in Fig. 25, the display output control unit 231 may control the display or output device 606 to display or output the BP time series information or the parameter adjustment time in the form shown in Fig. 26. Alternatively, or in addition to this, the display output control unit 231 may control to provide (display or output) the BP time series information or the parameter adjustment time to an operator, etc., as numerical information, date information, or time information. In this way, the information processing system 101 acquires already existing BP time series information, estimates BP time series information for a period after the present 2602 (future period, estimated period, subsequent period) based on the acquired BP time series information, and then determines the parameter adjustment time 2603 based on the estimated BP time series information, and controls to display or output the determined content. Alternatively, the information processing system 101 acquires already existing BP time series information, condition time series information, or adjustment time information, determines the parameter adjustment time 2603 for a period after the present 2602 (future period, estimated period, subsequent period) based on the acquired BP time series information, condition time series information, or adjustment time information, and controls to display or output the determined content. Therefore, the information processing system 101 can present an appropriate parameter adjustment time 2603 to an operator, etc. It becomes easier for the operator, etc. to make a plan for the adjustment of parameters for controlling the accelerator 170.

[0045] 5.Other (variations) The present disclosure is not limited to the above-described embodiment, but includes various modifications. A part of the configuration or processing of the embodiment may be replaced with the configuration or processing of another conceivable embodiment. The configuration or processing of the embodiment may be added to the configuration or processing of another conceivable embodiment.

[0046] For example, the present disclosure may have the following modified embodiments.

[0047] (α) Selective Execution In the flowchart of FIG. 7 in the above embodiment, when the BP (or deviation 301) for each BPM deviates from the adjustment-free range 302, a cause estimation process in step 704, a parameter set adjustment (creation) process in step 706 (if the judgment result in step 705 is positive), and a warm-up operation judgment process in step 709 are executed. However, some of these processes may not be performed. For example, the cause estimation process of step 704 may not be performed, and the parameter set adjustment (creation) process of step 706 may be performed. This modified example makes it possible to simplify the processing performed by the information processing system 101 in cases where it is not necessary to place emphasis on the cause of the deviation of the BP (or deviation 301) for each BPM from the adjustment-free range 302. In addition, in the case where the cause estimation process of step 704 is not performed, as a conditional judgment of whether or not step 706 is performed, a conditional branching process similar to step 802 in Fig. 8 may be performed instead of step 705. That is, when the BPs (or beam trajectories 173) in all BPMs are on the same side as seen from the design beam trajectory (when seen from the center of the accelerator 170, all the beam trajectories 173 are on the outside or inside of the design beam trajectory) (or, to simplify the conditional judgment, in the case of the accelerator 170 as shown in Fig. 1 or Fig. 2, it may be judged that the BPs (or beam trajectories 173) in the BPMs 181 and 183, which are located opposite each other, are on the same side as seen from the design beam trajectory (all on the outside or all on the inside).), the parameter set adjustment (creation) process of step 706 may be performed. With such a modified example, it is possible to perform the parameter set adjustment (creation) process only in cases where it is relatively easy to deal with the deviation of the beam orbit 173 (for example, a case where it is possible to correct the deviation of the beam orbit 173 from the designed beam orbit by only adjusting the deflection component of the magnetic field components (deflection component and steering component) that the deflection electromagnet with steering function applies to the circular path 172). In other words, it is possible to simplify the parameter set adjustment (creation) process performed by the information processing system 101. Also, for example, the parameter set adjustment (creation) process of step 706 may be executed while the warm-up operation determination process of step 711 (or the cooling operation determination process described later) may not be executed. Conversely, the parameter set adjustment (creation) process of step 706 may not be executed while the warm-up operation determination process of step 711 (or the cooling operation determination process described later) may be executed. This modification can simplify the processing performed by the information processing system 101 in cases where it is sufficient to perform either the parameter set adjustment (creation) or the warm-up operation determination (or the cooling operation determination described below).

[0048] (β) Selective use of information (ingredients, items) In the above embodiment, information on the 14 components (items) shown in FIG. 4 and causes of the 7 components (items) shown in FIG. 5 are handled. However, it is not necessary to use all of the information and causes of these components (items) to perform the cause estimation process, the parameter adjustment (creation) process, the warm-up operation determination process, and the adjustment timing determination process. Only a part of the information and causes of these components (items) may be used. In addition, the cause estimation process, parameter adjustment (creation) process, warm-up operation determination process, and adjustment timing determination process may be performed using information or causes other than the components (items) discussed in the above embodiment. In this way, the present disclosure makes it possible to flexibly define how information is handled so that various processes can be performed according to the circumstances of each accelerator.

[0049] (γ) Utilizing information from previous parameter adjustments In the above embodiment, information on the time when the parameters for controlling the accelerator 170 were previously adjusted and information on the time elapsed since the parameters for controlling the accelerator 170 were most recently adjusted, which is derived from the information indicated by the components (items)

[11] in FIG. 4, may be used in the adjustment timing determination process (e.g., the process shown in [FIGS. 25] to [FIGS. 31]). Information on the time when the parameters for controlling the accelerator 170 were previously adjusted and information on the time elapsed since the parameters for controlling the accelerator 170 were most recently adjusted, which are derived from the information indicated by the component (item)

[11] shown in Fig. 4, may also be used in each of the cause estimation process (e.g., the process shown in [Fig. 8] to [Fig. 12]), the parameter adjustment (creation) process (e.g., the process shown in [Fig. 13] to [Fig. 18]), and the warm-up operation determination process (e.g., the process shown in [Fig. 20] to [Fig. 24], including the cooling operation determination process in the modified example shown later). When used in this way, the information indicated by the component

[11] shown in Fig. 4 may be treated as one of the condition information. In this way, the present disclosure can perform processing that further incorporates information on the operating history of the accelerator in each of the cause estimation processing, parameter adjustment (creation) processing, and warm-up operation judgment processing (including the cooling operation judgment processing in the modified example shown later).

[0050] (δ) Relaxation of the criteria in the orbit comparison step In the above embodiment, in step 802, which may be considered as an trajectory comparison step, the cause estimation unit 221 examines the positional relationship between the design beam trajectory and the beam trajectory 173, and determines whether the situation is such that for any of the BPMs 181-184 (or a pair of BPMs in opposing positions), the beam trajectory 173 is located outside the design beam trajectory (as viewed from the center of the accelerator 170), or whether the beam trajectory 173 is located inside the design beam trajectory (as viewed from the center of the accelerator 170) for any of the BPMs 181-184 (or a pair of BPMs in opposing positions). Furthermore, in the above description of the embodiment and the above description of (α), it is also indicated that instead of the determination process shown as step 704, a determination process similar to step 802 can be performed. It is expected that the adjustment (creation) of the parameters (set) for correcting the deviation of the beam trajectory 173 from the design beam trajectory can be automated to a certain extent, even when the judgment result of step 802 is positive, when the judgment result of step 704 is positive, or when the judgment result of the judgment process in the modified example in which the same judgment process as step 802 is performed instead of the judgment process shown as step 704 is positive. Therefore, when the judgment result of the above-mentioned judgment process is positive, in step 706 which may be a parameter acquisition step, the parameter acquisition calculation unit 222 performs the adjustment (creation) process of the parameters (set), and in step 707 which may be a parameter display output step, the display output control unit 231 controls so as to display or output the adjusted parameters (set). Here, the parameter acquisition step and the parameter display output step may be performed when it is expected that the adjustment (creation) of parameters (sets) for correcting the deviation of the beam trajectory 173 from the design beam trajectory can be automated to a certain extent, and therefore the criteria for the judgment result in the above judgment process to be positive may be relaxed. For example, the judgment result in the above-mentioned judgment process may be made positive (the parameter acquisition step or the parameter display output step may be executed) not only when the strict condition that the position (deviation) of the beam orbit 173 from the design beam orbit (as viewed from the center of the accelerator 170) is on the same side (all outside or all inside) is satisfied in all BPMs targeted for checking the positional relationship between the design beam orbit and the beam orbit 173, but also when the relaxed condition that the position (deviation) of the beam orbit 173 from the design beam orbit (as viewed from the center of the accelerator 170) is on the same side is satisfied in a predetermined percentage of the BPMs (for example, 75% of the BPMs) among the BPMs targeted for checking the positional relationship between the design beam orbit and the beam orbit 173. Alternatively, for example, not only when the strict conditions described above are satisfied, but also when a relaxed condition is satisfied that even if there are BPMs in which the position (deviation) of the beam orbit 173 from the design beam orbit (seen from the center of the accelerator 170) is on a different side, the absolute value of the deviation 301 at the BPM in which the deviation to one side is observed is equal to or less than a threshold value (for example, within a range 302 in which adjustment is not required), the judgment result in the above-mentioned judgment process may be deemed positive (the parameter acquisition step or the parameter display output step may be executed). As described above, for the above-mentioned judgment process, it is possible to set rational conditions that are expected to enable a certain degree of automation in the adjustment (creation) of parameters (sets) for correcting the deviation of the beam trajectory 173 from the design beam trajectory, thereby increasing the opportunities for adjusting (creating) the parameters (sets). Depending on the operational circumstances of the accelerator 170, the conditions for allowing the parameter acquisition step and the parameter display output step to be executed may be set to be stricter, contrary to the above.

[0051] (ε) Decision regarding cooling operation (cooling operation decision process) (modification of C) As shown in "4.1.5. Determination regarding warm-up operation (C)," in the above embodiment, the warm-up operation determination unit 223 determines whether or not to perform warm-up operation in the accelerator 170, and if warm-up operation is performed, determines what form the warm-up operation should take. As already pointed out, the particle beam therapy system 160 (e.g., the accelerator 170) may be determined not only for warm-up operation but also for cooling operation. It is assumed that a change in the temperature of the accelerator (e.g., the temperature of an electromagnet used in the accelerator) may affect the beam trajectory of the accelerator. Here, not only a temperature lower than expected but also a temperature higher than expected may cause an operational problem of the accelerator. To deal with this, the particle beam therapy system 160 may be configured to be able to perform a warm-up operation or a cooling operation (including a temporary suspension of the particle beam therapy system 160). In a modified example for determining whether or not to perform cooling operation in the accelerator 170, and for determining the mode of cooling operation if cooling operation is performed, processing similar to the processing for warm-up operation shown in "4.1.5. Determination regarding warm-up operation (C)" may be performed for the cooling operation.

[0052] Specifically, in the modified example, a cooling operation determination unit may be provided as a functional unit included in the information processing system 101, or in the modified example, the warm-up operation determination unit 223 may also perform processing for determining the cooling operation. In the case where the warm-up operation determination unit 223 also performs processing for determining the cooling operation, this warm-up operation determination unit 223 may be called a "warm-up operation / cooling operation determination unit." In addition, steps similar to the steps related to the warm-up operation shown as steps 709 and 710 in Fig. 7 may be executed by the cooling operation determination unit or the warm-up operation / cooling operation determination unit. When both the determination related to the warm-up operation and the determination related to the cooling operation are performed in step 709, step 709 may be called a "warm-up operation / cooling operation determination step." When a determination regarding the cooling operation is made by a method similar to that shown in "4.1.5.1. Determination of warm-up operation based on cause estimation (C-1)", in a step corresponding to step 2002, the cooling operation determination unit or the warm-up operation cooling operation determination unit may make the determination result of the step corresponding to step 2002 positive if it is determined that the cooling water temperature is higher than a predetermined temperature threshold or that the elapsed time after the start-up or restart of the particle beam therapy system 160 is longer than a predetermined time threshold. If the determination result of the step corresponding to step 2002 is positive, in a step corresponding to step 2003, the cooling operation determination unit or the warm-up operation cooling operation determination unit may determine that the cooling operation is in progress as a result of the cooling operation determination process, and may appropriately determine the mode of the cooling operation (for example, the remaining time of the cooling operation (the duration from the current time) or the time zone of the cooling operation). When a judgment regarding cooling operation is made by a method similar to that shown in "4.1.5.2. Determination of warm-up operation by DB search (C-2)" or "4.1.5.3. Determination of warm-up operation using warm-up operation judgment model (C-3)", the component (item)

[13] included in record 401 shown in FIG. 4 may have information on the presence or absence of cooling operation (in addition to or instead of information on the presence or absence of warm-up operation), and the component (item)

[14] may have information on the mode of cooling operation (if cooling operation was performed) (for example, the remaining time of cooling operation (the duration from the date and time related to the record) or the time zone of cooling operation). In addition, instead of determining whether or not a cooling operation is being performed and the nature of the cooling operation, the cooling operation determination unit or the warm-up operation / cooling operation determination unit may determine whether or not a temporary suspension measure has been taken for the operation of the particle beam therapy system 160 or the accelerator 170 and the nature of the temporary suspension measure (e.g., the time for which the temporary suspension measure will continue, or the time period during which the temporary suspension measure will be implemented). As a variation of the above, the information processing system 101 may make a judgment as to whether to perform both warm-up operation and cooling operation, or may make a judgment as to whether to perform cooling operation but not whether to perform warm-up operation. According to the above modification, even if the temperature of the accelerator (for example, the temperature of the electromagnets used in the accelerator) is higher than expected, appropriate measures can be taken.

[0053] (ζ) Automatic setting 7 in the above embodiment, in step 712, the parameter setting unit 232 judges whether or not an instruction has been received from an operator, etc. to set the contents displayed or output in step 707 or step 711 (parameter set for the electromagnet or warm-up operation parameters), or contents further modified by the operator, etc., in the accelerator control device 240. In other words, the condition for performing the setting process in the accelerator control device 240 is that there is a clear instruction from the operator, etc. However, some or all of the setting process for the accelerator controller 240 may be executed automatically, without waiting for clear instructions from an operator or the like. For example, the parameter setting unit 232 may automatically set the adjusted parameter set created by the parameter set adjustment (creation) process of step 706 to the accelerator control device 240 (or store it in the parameter storage device 241) without requiring step 712. Alternatively, the parameter setting unit 232 may automatically set the warm-up operation parameters indicating the mode of the warm-up operation determined by the warm-up operation determination process of step 711 to the accelerator control device 240 without requiring step 712. In such a modified example, the step of the setting process for the accelerator controller 240 performed by the parameter setting unit 232 may be an automatic parameter setting step. With this type of modification, in cases where an operator does not need to check the parameter sets for the adjusted electromagnets or the recommended parameters for warm-up operation, settings can be made quickly in the accelerator control device 240.

[0054] (η) Remote monitoring In the above embodiment, steps related to display output, that is, steps 707, 708, and 711 in Fig. 7 and step 2503 in Fig. 25, are executed by the display output control unit 231 included in the information processing system 101. Step 712 related to receiving an instruction from an operator or the like is executed by the parameter setting unit 232 included in the information processing system 101. The information processing system 101 itself may be located in a location (local) close to the accelerator control device 240, or may be located in a remote location (remote) from the accelerator control device 240. However, if the information processing system 101 is located in a location (local) close to the accelerator control device 240, then both the steps related to display output and the steps related to receiving instructions from an operator, etc. will be performed in a location (local) close to the accelerator control device 240. However, even if the information processing system 101 is located (local) close to the accelerator control device 240, a system located remotely from the accelerator control device 240 may be involved in some way in some or all of the steps related to display output and the steps related to receiving instructions from an operator, etc.

[0055] FIG. 32 shows a modified example in which the information processing system 101 is located (local) close to the accelerator control device 240, and a system located remote from the accelerator control device 240 is somehow involved in some or all of the steps related to display output and the steps related to receiving instructions from an operator, etc. 32, the information processing system 101 has a communication control unit 3203. In addition, a remote system 3200 is capable of communicating with the information processing system 101 via a network 3201. The remote system 3200 has a remote display output control unit 3231 and a remote setting instruction unit 3232. When steps corresponding to steps 707, 708, and 711 in FIG. 7 and step 2503 in FIG. 25, which are steps related to display output, are to be executed, the communication control unit 3203 in the information processing system 101 transmits information to be displayed to the remote (remote location) system 3200 via the network 3201. In the remote (remote location) system 3200 that receives the information, the remote display output control unit 3231 controls some display device or output device to display or output the information. Furthermore, the remote setting instruction unit 3232 in the remote (remote location) system 3200 accepts instructions from an operator or the like who has come into contact with the information (an instruction to set parameters for the electromagnet after adjustment and parameters for warm-up operation in the accelerator control device 240). The remote setting instruction unit 3232 transmits the contents of the accepted instruction to the information processing system 101 via the network 3201. In the information processing system 101 that has received the instruction, the parameter setting unit 232 performs setting processing on the accelerator control device 240 in accordance with the instruction. Furthermore, the parameter set for the adjusted electromagnet and the warm-up operation parameters that are controlled to be displayed or output by the remote display output control unit 3231 may be further modified by the remote (remote location) system 3200 (or an operator thereof), and the remote setting instruction unit 3232 may then receive an instruction to set the modified parameter (set) in the accelerator control device 240. In such a modified example, the step of the process performed by the communication control unit 3203 may be a communication control step. Also, in such a modified example, the step of the setting process for the accelerator control device 240 performed by the parameter setting unit 232 may be a parameter remote post-instruction setting step. With this modification, even if the information processing system 101 is located (local) close to the accelerator control device 240, it is possible to monitor the accelerator 170 and give setting instructions to the accelerator control device 240 at a remote location (remote) from the accelerator control device 240. This allows the number of operators, etc., located at the local locations where the accelerators 170 are installed to be reduced and centralized management at the remote locations (remote), thereby enabling effective use of human resources such as operators. In the above-described modified example, the remote (remote location) system 3200 may have the remote display output control unit 3231 but may not have the remote setting instruction unit 3232. In other words, the remote (remote location) system 3200 may control the display and output of the results of each process, but may not instruct the accelerator control device 240 to set parameters. In this case, the information processing system 101 may instruct the accelerator control device 240 to set parameters, or the parameters may be automatically set in the accelerator control device 240. In this manner, the relationship between the accelerator control device 240, the information processing system 101, and the remote system 3200 can be flexibly determined.

[0056] (θ) Tuning model parameters for multiple records in machine learning In the above embodiment, in steps 1105 and 1106 in Figure 11, steps 1605 and 1606 in Figure 16, steps 2305 and 2306 in Figure 23, and steps 3005 and 3006 in Figure 30, the error (Loss function) is calculated for each record used in the machine learning of various models (in Figure 30, for each group of records included in a pair of preceding and succeeding periods), and then the model parameters are adjusted. However, in a modified example related to machine learning, the model parameters may be adjusted after calculating an overall error (Loss function) for a group of records to be treated collectively (a group of records corresponding to a group of pairs in FIG. 30) in a batch of multiple records (multiple pairs of a preceding period and a following period in FIG. 30) or in a batch of all records used in machine learning (all pairs of a preceding period and a following period in FIG. 30). The above-described modified example can reduce the frequency of adjustment of model parameters and can perform learning processing in accordance with the standard characteristics exhibited by the set of records used in machine learning.

[0057] The technical matters described in each of the embodiments of the present disclosure and the modified examples of the embodiments described above can be combined as appropriate as long as no technical contradiction occurs.

Claims

1. 1. A method for beam monitoring in a particle beam therapy system, comprising: The particle beam therapy system includes an accelerator for accelerating a charged particle beam; the accelerator includes a circulation path for circulating the charged particle beam, a plurality of beam monitors arranged in sequence along the circulation path, and a plurality of electromagnets arranged in sequence along the circulation path, the beam monitor acquires information on a beam position, which is a position of the charged particle beam when the charged particle beam passes near the beam monitor during rotation; the electromagnet adjusts a beam trajectory of the charged particle beam when the circulating charged particle beam passes near the electromagnet; The beam monitoring method is executed by an information processing system, The beam monitoring method includes: a beam position information acquisition step of acquiring information on the beam position in each of the beam monitors; A beam monitoring method comprising a parameter acquisition step of acquiring parameters to be set for each of the electromagnets, which can be used to adjust the beam trajectory of the orbiting charged particle beam, based on the beam position information acquired by the beam position information acquisition step.

2. 2. A method for monitoring a beam as claimed in claim 1, comprising the steps of: The parameter acquisition step is a step of calculating the parameters to be set for each electromagnet so that the beam orbit of the charged particle beam in orbit approaches the designed beam orbit by calculating a Closed Orbit Distortion (COD) correction that adjusts the amount of magnetic field excitation applied to the orbital path by the electromagnet based on the deviation between the beam position information in each beam monitor and a normal value corresponding to a designed beam orbit.

3. 2. A method for monitoring a beam as claimed in claim 1, comprising the steps of: The beam monitoring method includes a condition information acquisition step of acquiring, as condition information, at least one of information among an outside air temperature, a room temperature of a room housing the accelerator, a date or time, an elapsed time after startup or restart of the particle beam therapy system, a temperature of cooling water, an elapsed time since construction of a building housing the accelerator, a position of an installation surface of the accelerator, parameters set in the electromagnet, and information on a time when parameters to be set in the electromagnet were previously adjusted; A beam monitoring method, wherein the parameter acquisition step is a step of acquiring parameters to be set in the electromagnet based on information on the beam position in each of the beam monitors or the acquired condition information.

4. 2. A method for monitoring a beam as claimed in claim 1, comprising the steps of: the accelerator includes, as the electromagnet, a plurality of bending electromagnets arranged in sequence along the circular path; The beam monitoring method includes a trajectory comparison step of examining a positional relationship between a beam trajectory of the charged particle beam orbiting near each of the beam monitors and a designed beam trajectory; a parameter acquisition step being executed when the orbit comparison step determines that, in the vicinity of any of the beam monitors, the beam orbit of the orbiting charged particle beam is located outside the designed beam orbit, or that, in the vicinity of any of the beam monitors, the beam orbit of the orbiting charged particle beam is located inside the designed beam orbit.

5. 2. A method for monitoring a beam as claimed in claim 1, comprising the steps of: the accelerator includes, as the electromagnets, a first bending electromagnet, a second bending electromagnet, a third bending electromagnet, and a fourth bending electromagnet that are installed in order along the circular path; the accelerator includes a first beam monitor as the beam monitor between the first bending electromagnet and the second bending electromagnet, the accelerator includes a second beam monitor as the beam monitor between the third bending electromagnet and the fourth bending electromagnet, The beam monitoring method includes an orbit comparison step of comparing information of the beam position in the first beam monitor with a normal value corresponding to a designed beam orbit, and comparing information of the beam position in the second beam monitor with a normal value corresponding to the designed beam orbit, the parameter acquisition step is executed when it is found by the trajectory comparison step that the beam trajectory of the circulating charged particle beam is located outside the designed beam trajectory both in the vicinity of the first beam monitor and in the vicinity of the second beam monitor, or that the beam trajectory of the circulating charged particle beam is located inside the designed beam trajectory both in the vicinity of the first beam monitor and in the vicinity of the second beam monitor, The beam monitoring method includes a parameter display output step of displaying or outputting the parameters acquired in the parameter acquisition step when the trajectory comparison step determines that the beam trajectory of the orbiting charged particle beam is located outside the designed beam trajectory both in the vicinity of the first beam monitor and in the vicinity of the second beam monitor, or that the beam trajectory of the orbiting charged particle beam is located inside the designed beam trajectory both in the vicinity of the first beam monitor and in the vicinity of the second beam monitor.

6. 2. A method for monitoring a beam as claimed in claim 1, comprising the steps of: a condition information acquisition step of acquiring, as condition information, at least one of the following information: outside air temperature, room temperature of a room in which the accelerator is housed, date or time, time elapsed since start-up or restart of the particle beam therapy system, temperature of cooling water, time elapsed since construction of a building in which the accelerator is housed, position of an installation surface of the accelerator, parameters set in the electromagnet, and information on a time when parameters to be set in the electromagnet were previously adjusted; a cause estimation step of estimating, when a deviation of the beam orbit of the orbit of the charged particle beam in orbit falls outside a range in which adjustment is not required, based on the beam position information in each of the beam monitors or the acquired condition information, inferring a cause of the beam orbit of the orbit of the charged particle beam in orbit falling outside a range in which adjustment is not required, in any of the beam monitors.

7. 7. A method for monitoring a beam as claimed in claim 6, comprising the steps of: A beam monitoring method comprising a warm-up operation / cooling operation determination step for obtaining a determination result that it is desirable to perform a warm-up operation or a cooling operation, instead of acquiring the parameters by the parameter acquisition step, or in addition to acquiring the parameters by the parameter acquisition step, when the causes estimated in the cause estimation step include either the elapsed time after startup or restart of the particle beam therapy system or the temperature of the cooling water.

8. 2. A method for monitoring a beam as claimed in claim 1, comprising the steps of: a condition information acquisition step of acquiring, as condition information, at least one of the following information: outside air temperature, room temperature of a room in which the accelerator is housed, date or time, time elapsed since start-up or restart of the particle beam therapy system, temperature of cooling water, time elapsed since construction of a building in which the accelerator is housed, position of an installation surface of the accelerator, parameters set in the electromagnet, and information on a time when parameters to be set in the electromagnet were previously adjusted; A beam monitoring method comprising a warm-up operation / cooling operation determination step of determining whether or not warm-up operation or cooling operation is necessary, or the type of warm-up operation or cooling operation to be performed, based on information on the beam position in each of the beam monitors or the acquired condition information.

9. 2. A method for monitoring a beam as claimed in claim 1, comprising the steps of: a beam position time series information acquisition step of acquiring time series information of the beam position in the beam monitor; a beam position time series information estimation step of estimating time series information of the beam position information in a future period based on the acquired time series information of the beam position information; A beam monitoring method comprising an adjustment timing determination step of determining when to adjust parameters to be set for the electromagnet based on time series information of the estimated beam position information for a future period.

10. 2. A method for monitoring a beam as claimed in claim 1, comprising the steps of: a beam position time series information acquisition step of acquiring time series information of the beam position in the beam monitor; a condition time-series information acquisition step of acquiring, as condition time-series information, time-series information consisting of at least one of information on an outside air temperature, a room temperature in a room in which the accelerator is housed, a date or time, an elapsed time after the start-up or restart of the particle beam therapy system, a temperature of a cooling water, an elapsed time after the construction of a building in which the accelerator is housed, a position of an installation surface of the accelerator, and a parameter set in the electromagnet; an adjustment time information acquisition step of acquiring information on a time when a parameter to be set in the electromagnet was previously adjusted as adjustment time information; A beam monitoring method comprising an adjustment timing determination step of determining when to adjust parameters to be set in the electromagnet based on the acquired time series information of the beam position, the acquired condition time series information, and the acquired adjustment timing information.

11. 2. A method for monitoring a beam as claimed in claim 1, comprising the steps of: A beam monitoring method comprising an automatic parameter setting step of transferring and storing the acquired parameters in a storage device for an accelerator control device that controls the accelerator, without receiving an input indicating a setting instruction regarding the parameters acquired by the parameter acquisition step.

12. 2. A method for monitoring a beam as claimed in claim 1, comprising the steps of: a parameter display output step of displaying or outputting the parameters acquired by the parameter acquisition step; A beam monitoring method comprising a parameter instruction post-setting step of transferring and storing the acquired parameters or further modified parameters in a storage device for an accelerator control device that controls the accelerator, on the condition that an input indicating a setting instruction regarding the displayed or output parameters is received.

13. 2. A method for monitoring a beam as claimed in claim 1, comprising the steps of: a communication control step of transmitting the parameters acquired by the parameter acquisition step to a system at a remote location from a building in which the accelerator is installed, so as to display or output the parameters acquired by the parameter acquisition step to the system at the remote location; A beam monitoring method comprising a parameter remote instruction post-setting step of transferring and storing the parameters transmitted by the communication control step or further modified parameters in a storage device for an accelerator control device that controls the accelerator, on the condition that a setting instruction regarding the parameters is received from the remote system.

14. A beam monitoring program for an information processing system for monitoring a particle beam therapy system, comprising: The particle beam therapy system includes an accelerator for accelerating a charged particle beam; the accelerator includes a circulation path for circulating the charged particle beam, a plurality of beam monitors arranged in sequence along the circulation path, and a plurality of electromagnets arranged in sequence along the circulation path, the beam monitor acquires information on a beam position, which is a position of the charged particle beam when the charged particle beam passes near the beam monitor during rotation; the electromagnet adjusts a beam trajectory of the charged particle beam when the circulating charged particle beam passes near the electromagnet; The beam monitoring program is provided to the information processing system. a beam position information acquisition step of acquiring information on the beam position in each of the beam monitors; A beam monitoring program for executing a parameter acquisition step for acquiring parameters to be set for each of the electromagnets, which can be used to adjust the beam orbit of the orbiting charged particle beam, based on the beam position information acquired by the beam position information acquisition step.

15. An information processing system for monitoring a particle beam therapy system, comprising: The particle beam therapy system includes an accelerator for accelerating a charged particle beam; the accelerator includes a circulation path for circulating the charged particle beam, a plurality of beam monitors arranged in sequence along the circulation path, and a plurality of electromagnets arranged in sequence along the circulation path, the beam monitor acquires information on a beam position, which is a position of the charged particle beam when the charged particle beam passes near the beam monitor during rotation; the electromagnet adjusts a beam trajectory of the charged particle beam when the circulating charged particle beam passes near the electromagnet; The information processing system includes: a beam position information acquisition unit that acquires information on the beam position in each of the beam monitors; an information processing system comprising: a parameter acquisition calculation unit that acquires parameters to be set for each of the electromagnets, which can be used to adjust the beam trajectory of the circulating charged particle beam, based on the beam position information acquired by the beam position information acquisition unit.

16. 1. A particle beam system comprising: The particle beam system includes a particle beam therapy system and an information processing system for monitoring the particle beam therapy system, The particle beam therapy system includes an accelerator for accelerating a charged particle beam; the accelerator includes a circulation path for circulating the charged particle beam, a plurality of beam monitors arranged in sequence along the circulation path, and a plurality of electromagnets arranged in sequence along the circulation path, the beam monitor acquires information on a beam position, which is a position of the charged particle beam when the charged particle beam passes near the beam monitor during rotation; the electromagnet adjusts a beam trajectory of the charged particle beam when the circulating charged particle beam passes near the electromagnet; The information processing system includes: a beam position information acquisition unit that acquires information on the beam position in each of the beam monitors; A particle beam system comprising a parameter acquisition calculation unit that acquires parameters to be set for each of the electromagnets, which can be used to adjust the beam trajectory of the circulating charged particle beam, based on the beam position information acquired by the beam position information acquisition unit.

Images