Information processing method, information processing device, and program

Abstract

This information processing device: acquires a setting condition; generates, on the basis of the setting condition, data in which a curve indicating the relationship between a plurality of parameters of the device is approximated using a polyline; and optimizes the plurality of parameters on the basis of the result of an evaluation performed when the device is operated using the data.

Description

Information processing method, information processing device, and program

[0001] The present disclosure relates to an information processing method, an information processing device, and a program.

[0002] Patent Literature 1 discloses a machine learning device according to background art, which includes a state observation unit that observes values ​​such as motor rotation speed detected by an electric motor control unit that drives a motor, a motor output calculation unit that calculates motor output from the values ​​such as motor rotation speed observed by the state observation unit, a reward calculation unit that calculates a reward based on the motor output, and a learning unit that updates an action value table based on values ​​such as motor rotation speed.

[0003] Optimizing servo motor parameters requires a huge amount of calculation, and there is a need to reduce the amount of calculation.

[0004] The machine learning device disclosed in Patent Document 1 does not consider reducing the amount of calculation required during learning.

[0005] Japanese Patent Application Laid-Open No. 2017-70105

[0006] An object of the present disclosure is to provide an information processing method, an information processing device, and a program that can significantly reduce the amount of calculation required for parameter optimization, thereby significantly shortening the time required for optimization.

[0007] In an information processing method according to one aspect of the present disclosure, an information processing device acquires setting conditions, generates data that approximates a curve showing the relationship between multiple parameters of the device using broken lines based on the setting conditions, and optimizes the multiple parameters based on evaluation results when the device is operated using the data.

[0008] FIG. 1 is a simplified diagram illustrating an overall configuration of a production system according to an embodiment of the present disclosure; FIG. 2 is a diagram illustrating a broken line characteristic defined by sequence data; FIG. 3 is a diagram illustrating a first example of sequence data; FIG. 4 is a diagram illustrating a second example of sequence data; FIG. 5 is a flowchart illustrating a learning phase process performed by a processing unit of an optimization device; FIG. 6 is a simplified diagram illustrating an example of a setting screen displayed on a display unit; FIG. 7 is a diagram illustrating a plurality of broken line characteristics that are gradually optimized; and FIG. 8 is a flowchart illustrating a utilization phase process performed by a processing unit of a setting device.

[0009] (Knowledge forming the basis of the present disclosure) In many production sites, servo motors are used as drive sources for driving production equipment such as mounting machines with high precision. In order to drive production equipment with high precision using servo motors, it is necessary to appropriately set parameters of the servo motors depending on the performance or operating conditions of the servo motors deployed at each production site.

[0010] However, since there are multiple parameters to be set for a servo motor and these parameters affect each other, when the value of the first parameter is changed, the optimal value of the second parameter also changes accordingly, and therefore the relationship between the optimal values ​​of the first parameter and the second parameter is expressed by a curve.

[0011] Because there are countless points on a curve, optimizing multiple parameters whose optimal value relationships are expressed by a curve using machine learning requires a huge amount of calculations and takes a long time to learn. Since it is not realistic to perform brute force calculations for the countless points on the curve, the operating conditions or search range, etc. are limited by utilizing the domain knowledge of experienced engineers. However, the amount of calculation is still enormous, and there is a need to reduce it.

[0012] In order to solve this problem, the present inventor discovered that the amount of calculations can be significantly reduced by approximately representing a curve showing the relationship between multiple parameters with a broken line and obtaining an approximate solution rather than an exact solution, and thus came up with the present disclosure.

[0013] Next, each aspect of the present disclosure will be described.

[0014] An information processing method according to a first aspect of the present disclosure is an information processing method for optimizing parameters of a device, in which an information processing device acquires setting conditions, generates data that approximates a curve showing the relationship between multiple parameters of the device using broken lines based on the setting conditions, and optimizes the multiple parameters based on evaluation results when the device is operated using the data.

[0015] According to the first aspect, data is generated that approximates a curve showing the relationship between multiple parameters using broken lines, and the multiple parameters are optimized based on the evaluation results when the device is operated using the data. By obtaining an approximate solution using broken lines rather than an exact solution using curves, the amount of calculation can be significantly reduced, and therefore the time required for optimization can be significantly shortened.

[0016] In an information processing method according to a second aspect of the present disclosure, in the first aspect, the plurality of parameters include a first parameter and a second parameter, the setting conditions include settings of the first parameter and the second parameter, and in generating the data, sequence data is generated that represents the curve defined on a plane including a first axis corresponding to the first parameter and a second axis corresponding to the second parameter as a sequence of change rate data indicating a change rate in the second axis direction in unit sections obtained by dividing the first axis at a predetermined interval.

[0017] According to the second aspect, by expressing a curve showing the relationship between a first parameter and a second parameter as a sequence of change rate data showing the change rate in a unit interval, a broken line that approximates the curve can be easily generated.

[0018] In the information processing method according to the third aspect of the present disclosure, in the second aspect, in optimizing the plurality of parameters, it is preferable to optimize the plurality of parameters by updating the sequence data based on evaluation results when operating the device using the sequence data.

[0019] According to the third aspect, since a plurality of parameters are optimized by updating sequence data including discrete degree-of-change data, the amount of calculation can be significantly reduced.

[0020] In the information processing method according to a fourth aspect of the present disclosure, in the third aspect, a metaheuristic search algorithm may be used to update the sequence data.

[0021] According to the fourth aspect, the amount of calculation can be further reduced by using a metaheuristic search algorithm, which is an approximate solution method.

[0022] In the information processing method according to the fifth aspect of the present disclosure, in the fourth aspect, the metaheuristic search algorithm may include a genetic algorithm.

[0023] According to the fifth aspect, the sequence data can be appropriately updated by using a genetic algorithm.

[0024] In the information processing method according to the sixth aspect of the present disclosure, in any one of the second to fifth aspects, the setting conditions may further include setting the number of unit intervals, and the number of unit intervals may be set according to the performance of the device or the operating conditions of the device.

[0025] According to the sixth aspect, the number of unit sections is set according to the performance or operating conditions of the device, so that optimal parameters tailored to each production site to which the device is applied can be output. As a result, when the device is applied to each production site, re-optimization at each production site is not required.

[0026] In the information processing method according to the seventh aspect of the present disclosure, in any one of the second to sixth aspects, the setting conditions may further include setting the predetermined interval, and the predetermined interval may be set according to the performance of the device or the operating conditions of the device.

[0027] According to the seventh aspect, the predetermined interval is set according to the performance or operating conditions of the device, so that optimal parameters tailored to each production site to which the device is applied can be output, eliminating the need for re-optimization at each production site when the device is applied to each production site.

[0028] In the information processing method according to the eighth aspect of the present disclosure, in any one of the second to seventh aspects, the setting conditions may further include setting the number of bits of the change rate data, and the number of bits of the change rate data may be set according to the performance of the device or the operating conditions of the device.

[0029] According to the eighth aspect, the number of bits of the change rate data is set according to the performance or operating conditions of the device, so that optimal parameters tailored to each production site to which the device is applied can be output, eliminating the need for re-optimization at each production site when the device is applied to each production site.

[0030] In the information processing method according to a ninth aspect of the present disclosure, in any one of the first to eighth aspects, the operation of the device may include an operation by simulation.

[0031] According to the ninth aspect, by optimizing a plurality of parameters based on the evaluation results during operation by simulation, the time required for optimization can be further reduced.

[0032] In an information processing method according to a tenth aspect of the present disclosure, in any one of the first to ninth aspects, when acquiring the setting conditions, it is preferable to display a setting screen including an input field for the setting conditions, and acquire the setting conditions entered in the input field by a user's input operation.

[0033] According to the tenth aspect, a setting screen including input fields for setting conditions is displayed, and the user is prompted to perform input operations, thereby making it possible to easily obtain setting conditions suited to each production site.

[0034] An information processing method according to an eleventh aspect of the present disclosure is any one of the first to tenth aspects, wherein the device includes a servo motor.

[0035] According to the eleventh aspect, the amount of calculation required to obtain the optimum parameters of the servo motor can be significantly reduced.

[0036] An information processing device according to a twelfth aspect of the present disclosure is an information processing device that optimizes device parameters, acquires setting conditions, generates data that approximates a curve showing the relationship between multiple parameters of the device using broken lines based on the setting conditions, and optimizes the multiple parameters based on evaluation results when the device is operated using the data.

[0037] According to the twelfth aspect, data is generated in which a curve showing the relationship between multiple parameters is approximated by a broken line, and the multiple parameters are optimized based on the evaluation results when the device is operated using the data. By obtaining an approximate solution using a broken line rather than an exact solution using a curve, the amount of calculation can be significantly reduced, and therefore the time required for optimization can be significantly shortened.

[0038] A program according to a thirteenth aspect of the present disclosure is a program for causing an information processing device to execute a process that optimizes device parameters, the process acquiring setting conditions, generating data that approximates a curve showing the relationship between multiple parameters of the device using broken lines based on the setting conditions, and optimizing the multiple parameters based on evaluation results when the device is operated using the data.

[0039] According to the thirteenth aspect, data is generated in which a curve showing the relationship between a plurality of parameters is approximated by a broken line, and the plurality of parameters are optimized based on the evaluation results when the device is operated using the data. By obtaining an approximate solution using a broken line rather than an exact solution using a curve, the amount of calculation can be significantly reduced, and therefore the time required for optimization can be significantly shortened.

[0040] The present disclosure can also be realized as a program that causes a computer to execute each characteristic configuration included in such a method or apparatus, or as a system operated by this program. Needless to say, such a computer program can be distributed on a computer-readable non-transitory recording medium such as a CD-ROM or via a communication network such as the Internet.

[0041] (Embodiments of the Present Disclosure) Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. Elements with the same reference numerals in different drawings indicate the same or corresponding elements. Furthermore, the components, the arrangement positions of the components, the connection forms, the order of operations, etc. shown in the following embodiments are merely examples and are not intended to limit the present disclosure. The present disclosure is limited only by the claims. Therefore, among the components in the following embodiments, components that are not described in the independent claims that represent the highest concept of the present disclosure are not necessarily required to achieve the objectives of the present disclosure, but are described as constituting more preferred forms.

[0042] FIG. 1 is a simplified diagram showing the overall configuration of a production system according to an embodiment of the present disclosure. The production system includes an optimization device 1, a setting device 2, and a servo system 3. The optimization device 1 and the setting device 2 may be personal computers or server devices. The optimization device 1 and the setting device 2 may be configured as a single terminal that shares hardware. The optimization device 1, the setting device 2, and the servo system 3 can communicate with each other via a communication network NW. The communication network NW may be a dedicated line network or a public line network.

[0043] The servo system 3 is installed at a production site. Although only one servo system 3 is shown in FIG. 1, in reality, there are multiple production sites, and a servo system 3 is installed at each production site. The servo system 3 has a servo motor 31 and a driven device 32. The driven device 33 is any production equipment such as a mounting machine, and is driven by the servo motor 31. Since the servo system 3 installed differs at each production site, the optimal parameters of the servo motor 31 also differ at each production site. In other words, the optimal parameters of the servo motor 31 differ depending on the performance or operating conditions of the servo motor 31 of the servo system 3.

[0044] The optimization device 1 includes a processing unit 11, a memory unit 12, an input unit 13, a display unit 14, and a communication unit 15. The setting device 2 includes a processing unit 21, a memory unit 22, and a communication unit 23. The processing units 11 and 21 include a processor such as a CPU. The memory units 12 and 22 include a HDD, an SSD, or a semiconductor memory. The input unit 13 includes an input device such as a keyboard, a mouse, or a touch panel. The display unit 14 includes an display device such as a liquid crystal display or an organic EL display. The communication units 15 and 23 include a communication module compatible with an IP or other communication standard.

[0045] The storage unit 12 stores a program 51, setting conditions 52, an updated model 53, sequence data 54, and optimal parameters 55. The storage unit 22 stores the optimal parameters 55. The storage unit 12 includes a computer-readable non-volatile storage medium, and the program 51 is stored in the storage medium.

[0046] The processing unit 11 has an acquisition unit 41, a generation unit 42, an operation execution unit 43, an evaluation unit 44, and an optimization unit 45 as functions realized by the processor executing the program 51 read from the storage unit 12. The processing content of each of these units will be described in detail later.

[0047] FIG. 2 is a diagram showing the broken-line characteristic K defined by the sequence data 54. To drive the driven device 33 with high precision using the servo motor 31, the parameters of the servo motor 31 must be appropriately set according to the performance or operating conditions of the servo motor 31 deployed at each production site. The parameters of the servo motor 31 include the number of rotations, rotational speed, torque, voltage, current, time, and output value. The servo motor 31 requires multiple parameters to be set, and these parameters (the first parameter P1 and the second parameter P2) influence each other. In other words, when the value of the first parameter P1 is changed, the optimal value of the second parameter P2 also changes accordingly. Therefore, the relationship between the first parameter P1 and the second parameter P2, which indicate the optimal values, is essentially expressed by the curve characteristic C. The curve characteristic C is defined on an XY plane including the X-axis (X≧0), which is the first axis corresponding to the first parameter P1, and the Y-axis (Y≧0), which is the second axis corresponding to the second parameter P2.

[0048] The broken line characteristic K is obtained by defining a plurality of points on the curve characteristic C at predetermined intervals W on the X-axis and connecting the plurality of points with straight lines. In other words, the broken line characteristic K corresponds to a characteristic expressed as a sequence of degree-of-change data indicating the degree of change in the Y-axis direction in each of a plurality of unit sections SC obtained by dividing the X-axis at predetermined intervals W. The degree of change may be an amount of change or a rate of change. In the following, an example will be described in which the amount of change B is used as the degree of change and the degree-of-change data is amount-of-change data. In the example of FIG. 2, the unit section SC includes six unit sections SC1 to SC6.

[0049] 3 is a diagram showing a first example of the sequence data 54. The sequence data 54 is a sequence of a plurality of change amount data. Each change amount data indicates a change amount B in the Y-axis direction for each unit section SC1 to SC6. In the first example, the maximum value of the change amount B is "12" in decimal notation, and since it is necessary to distinguish between positive and negative signs, the number of data bits for each change amount data expressed in binary notation is 5.

[0050] 4 is a diagram showing a second example of the sequence data 54. In the second example, the value of the change amount data is "0" when the change amount B is positive, "2" when the change amount B is negative, and "1" when the change amount B is zero. In the second example, the number of data bits of each change amount data expressed in binary is 2 bits, and the data amount of the sequence data 54 is reduced compared to the first example.

[0051] FIG. 5 is a flowchart showing the processing of the learning phase executed by the processing unit 11 of the optimization device 1.

[0052] First, in step SP11, the acquisition unit 41 acquires the setting conditions 52.

[0053] 6 is a simplified diagram showing an example of a setting screen 60 displayed on the display unit 14. The setting screen 60 has a setting item 61 related to the sequence data 54, a setting item 62 related to evaluation, a setting item 63 related to an optimization algorithm, and an icon 64 labeled "Start."

[0054] The setting item 61 has input fields 70 to 75. An initial value of the sequence data 54 is input into the input field 70 labeled "Initial Value." When operating the servo motor 31 from a stopped state, the initial value is set to "0." The number of unit intervals SC is input into the input field 71 labeled "Number of Intervals." The width of the unit interval SC is input into the input field 72 labeled "Interval Width." The width of the unit interval SC corresponds to the predetermined interval W. The number of data bits of each change amount data is input into the input field 73 labeled "Number of Data Bits." A parameter selected as the first parameter P1 is input into the input field 74 labeled "Parameter 1." A parameter selected as the second parameter P2 is input into the input field 75 labeled "Parameter 2." Note that adding input fields 74 and 75 may allow selection of three or more parameters.

[0055] The setting item 62 has input fields 76 and 77. An evaluation index, which is an objective function of optimization, is input into the input field 76 labeled "evaluation value." The evaluation index is, for example, settling time. A target value of the evaluation index is input into the input field 77 labeled "target value." If the evaluation index is settling time, the target value is the maximum allowable value of the settling time.

[0056] The setting item 63 has input fields 78 and 79. In the input field 78 labeled "algorithm," an algorithm for updating the sequence data 54 in the parameter optimization process is input. Algorithms that can be input include metaheuristic search algorithms. Metaheuristic search algorithms include genetic algorithms. Metaheuristic search algorithms may include tabu search, etc. The update model 53 stored in the memory unit 12 includes multiple update models corresponding to multiple algorithms that can be input. In the input field 79 labeled "number of generations," the number of generations of the initial population in a genetic algorithm, etc. is input.

[0057] The operator operates the input unit 13 to input information into the input fields 70 to 79 according to the performance or operating conditions of the servo motor 31 of the servo system 3 deployed at the production site. The performance includes performance limits. The performance limits include upper limits for rotation speed or torque, etc. The operating conditions include power supply voltage value, continuous operating time, reference rotation speed, control resolution, etc. The control resolution corresponds to the above-mentioned predetermined interval W. Note that the configuration may be such that limiting information such as operating conditions or search ranges can be set on the setting screen 60 to utilize the domain knowledge of experienced engineers.

[0058] After input of all the information is completed, when the operator selects icon 64 by operating input unit 13, acquisition unit 41 acquires the information input to setting screen 60 as setting conditions 52. Processing unit 11 stores setting conditions 52 in storage unit 12.

[0059] 5 , next, in step SP12, the generation unit 42 generates sequence data 54 based on the set conditions 52. When a genetic algorithm is input in the input field 78, the generation unit 42 generates a plurality of first-generation sequence data 54 corresponding to the number of generations input in the input field 79. The generation unit 42 randomly sets the value of the change amount data in each sequence data 54.

[0060] Next, in step SP13, the operation execution unit 43 operates the servo motor 31 by executing a simulation using each sequence data 54. Note that the operation execution unit 43 may actually operate the servo motor 31 using each sequence data 54 instead of or in addition to executing a simulation.

[0061] Next, in step SP14, the evaluation unit 44 calculates an evaluation value for each operation of the servo motor 31 using each sequence data 54.

[0062] Next, in step SP15, the evaluation unit 44 determines whether the best evaluation value among the plurality of evaluation values ​​corresponding to the plurality of sequence data 54 is equal to or greater than the target value.

[0063] If the best evaluation value is less than the target value (step SP15: NO), then in step SP16, the optimization unit 45 generates a second generation of multiple sequence data 54 by updating the best sequence data among the multiple sequence data 54 using the updated model 53 corresponding to the algorithm entered in the input field 78.

[0064] The processing unit 11 repeatedly executes the processes of steps SP13 to SP16 until the best evaluation value becomes equal to or greater than the target value, thereby gradually optimizing the parameters, and accordingly, the sequence data 54 and the broken line characteristic K.

[0065] 7 is a diagram showing a plurality of gradually optimized polygonal line characteristics. The best first-generation polygonal line characteristic K0 is updated to obtain the best second-generation polygonal line characteristic K1. The best second-generation polygonal line characteristic K1 is updated to obtain the best third-generation polygonal line characteristic K2. As the generations progress, the polygonal line characteristic K is gradually optimized.

[0066] If the best evaluation value is equal to or greater than the target value (step SP15: YES), the optimization unit 45 outputs the sequence data 54 and the optimal parameters 55 corresponding to the best evaluation value, and stores them in the storage unit 12.

[0067] FIG. 8 is a flowchart showing the process of the use phase executed by the processing unit 21 of the setting device 2.

[0068] First, in step SP21, the processing unit 21 acquires the optimal parameters 55 by receiving them from the optimization device 1. The processing unit 21 stores the acquired optimal parameters 55 in the storage unit 22. Note that the processing unit 21 may also receive the corresponding sequence data 54 and broken line characteristics K from the optimization device 1 along with the optimal parameters 55.

[0069] Next, in step SP22, the processing unit 21 transmits the optimum parameters 55 to the servo system 3, thereby setting the optimum parameters 55 for the servo motor 31. The servo motor 31 operates based on the optimum parameters 55 and drives the device 32 to be driven.

[0070] According to this embodiment, sequence data 54 is generated in which curves showing the relationships between multiple parameters are approximated by broken lines, and multiple parameters are optimized based on evaluation results when the servo motor 31 is operated using the sequence data 54. By obtaining an approximate solution using broken lines rather than an exact solution using a curve, the amount of calculation can be significantly reduced, and therefore the time required for optimization can be significantly shortened.

[0071] Furthermore, according to this embodiment, by expressing the curve characteristic C showing the relationship between the first parameter P1 and the second parameter P2 as a sequence of change amount data showing the change amount B for each unit section SC, it is possible to easily generate a broken line characteristic K that approximately represents the curve characteristic C.

[0072] Furthermore, according to this embodiment, since a plurality of parameters are optimized by updating the sequence data 54 including discrete change amount data, the amount of calculation can be significantly reduced.

[0073] Furthermore, according to this embodiment, the amount of calculation can be further reduced by using a metaheuristic search algorithm, which is an approximate solution method.

[0074] Furthermore, according to this embodiment, the sequence data 54 can be appropriately updated by using a genetic algorithm.

[0075] Furthermore, according to this embodiment, it is possible to output optimal parameters 55 tailored to each production site to which the system is applied, by setting the number of unit sections SC according to the performance or operating conditions of the servo motor 31. As a result, when the system is applied to each production site, re-optimization at each production site is not required.

[0076] Furthermore, according to this embodiment, it is possible to output the optimum parameters 55 suited to each production site to which the system is applied, by setting the predetermined interval W according to the performance or operating conditions of the servo motor 31. As a result, when the system is applied to each production site, there is no need to re-optimize the system at each production site.

[0077] Furthermore, according to this embodiment, it is possible to output optimal parameters 55 tailored to each production site to which the system is applied, by setting the number of data bits of the change amount data in accordance with the performance or operating conditions of the servo motor 31. As a result, when the system is applied to each production site, re-optimization at each production site is not required.

[0078] Furthermore, according to this embodiment, by optimizing a plurality of parameters based on the evaluation results during operation by simulation, the time required for optimization can be further reduced.

[0079] Furthermore, according to this embodiment, the setting screen 60 including the input fields 70 to 79 for the setting conditions 52 is displayed on the display unit 14, and the user is allowed to perform input operations, thereby easily obtaining the setting conditions 52 suited to each production site.

[0080] Furthermore, according to this embodiment, the servo motor 31 is included in the device whose parameters are to be optimized, so that the amount of calculation required to find the optimal parameters for the servo motor 31 can be significantly reduced.

[0081] The present disclosure is widely applicable to servo systems and the like that include a servo motor as a drive source.

Claims

1. An information processing method for optimizing parameters of a device, comprising: an information processing device acquires setting conditions; generates data that approximates a curve showing the relationship between multiple parameters of the device based on the setting conditions, using broken lines; and optimizes the multiple parameters based on evaluation results when the device is operated using the data.

2. The information processing method of claim 1, wherein the plurality of parameters include a first parameter and a second parameter, the setting conditions include settings of the first parameter and the second parameter, and in generating the data, sequence data is generated in which the curve defined on a plane including a first axis corresponding to the first parameter and a second axis corresponding to the second parameter is expressed as a sequence of change rate data indicating a change rate in the second axis direction in unit intervals obtained by dividing the first axis at a predetermined interval.

3. The information processing method according to claim 2, wherein in optimizing the plurality of parameters, the plurality of parameters are optimized by updating the sequence data based on evaluation results when the device is operated using the sequence data.

4. The information processing method according to claim 3, wherein a metaheuristic search algorithm is used to update the sequence data.

5. The information processing method according to claim 4, wherein the metaheuristic search algorithm includes a genetic algorithm.

6. The information processing method according to claim 2, wherein the setting conditions further include setting the number of unit intervals, and the number of unit intervals is set according to the performance of the device or an operating condition of the device.

7. The information processing method according to claim 2, wherein the setting conditions further include setting the predetermined interval, and the predetermined interval is set according to the performance of the device or an operating condition of the device.

8. The information processing method according to claim 2, wherein the setting conditions further include setting a number of bits of the degree of change data, and the number of bits of the degree of change data is set according to the performance of the device or an operating condition of the device.

9. The information processing method according to claim 1, wherein the operation of the device includes an operation by simulation.

10. The information processing method according to claim 1, wherein acquiring the setting conditions comprises displaying a setting screen including an input field for the setting conditions, and acquiring the setting conditions input into the input field by a user's input operation.

11. The information processing method of claim 1, wherein the device includes a servo motor.

12. An information processing device that optimizes device parameters, comprising: acquiring setting conditions; generating data that approximates a curve showing the relationship between multiple parameters of the device based on the setting conditions using broken lines; and optimizing the multiple parameters based on evaluation results when the device is operated using the data.

13. A program for causing an information processing device to execute a process for optimizing device parameters, the process comprising: acquiring setting conditions; generating data that approximates a curve showing the relationship between multiple parameters of the device using broken lines based on the setting conditions; and optimizing the multiple parameters based on evaluation results when the device is operated using the data.

Images