State assessment device, state assessment method, and state assessment program

Abstract

The present invention improves assessment accuracy in assessing the state of an air conditioner. A state assessment device (120) is for an air conditioner (110) including a monitored device and a non-monitored device, and comprises a control unit (400) that assesses the state of the air conditioner on the basis of first information which is a control amount or a measurement amount relating to the monitored device, and second information which is a control amount relating to the non-monitored device. The control unit (400) predicts the first information in a normal state on the basis of the second information acquired during operation of the air conditioner (110), and assesses the state of the air conditioner on the basis of the first information acquired during operation of the air conditioner (110) and the predicted first information in a normal state.

Description

State determination device, state determination method, and state determination program

[0001] The present disclosure relates to a state determination device, a state determination method, and a state determination program.

[0002] A known technique for determining the condition (abnormality or deterioration) of an air conditioner is to compare a measured quantity (actual value) measured by a specified sensor with a normal measured quantity (predicted value) of the specified sensor predicted based on a measured quantity (actual value) measured by a sensor other than the specified sensor.

[0003] According to the above technology, it is possible to determine whether an air conditioner is abnormal or deteriorated based on the degree of deviation between a measurement quantity (actual measurement value) measured by a specified sensor and a predicted measurement quantity (predicted value) under normal conditions.

[0004] Japanese Patent Application Laid-Open No. 2001-133011

[0005] However, with the above technology, in cases where an abnormality or the like occurs in an air conditioner and affects the measurement quantities (actual measured values) measured by sensors other than the specified sensor, the accuracy of determining whether an abnormality or the like in the air conditioner exists decreases. This is because the measurement quantities (predicted values) of the specified sensor under normal conditions, which are predicted based on the measurement quantities (actual measured values) measured by sensors other than the specified sensor, will deviate from the measurement quantities (actual measured values) measured by the specified sensor under normal conditions (i.e., the prediction accuracy decreases).

[0006] The present disclosure aims to improve the accuracy of determining the state of an air conditioner.

[0007] A first aspect of the present disclosure is a state determination device for an air conditioner including monitored equipment and non-monitored equipment, comprising a control unit that determines the state of the air conditioner based on first information which is a control quantity or measurement quantity related to the monitored equipment and second information which is a control quantity related to the non-monitored equipment, wherein the control unit predicts the first information in a normal state based on the second information acquired while the air conditioner is operating, and determines the state of the air conditioner based on the first information acquired while the air conditioner is operating and the predicted first information in a normal state.

[0008] According to the first aspect of the present disclosure, it is possible to improve the accuracy of determining the state of an air conditioner.

[0009] A second aspect of the present disclosure is the state determination device described in the first aspect, wherein the control unit acquires air conditioner information while the air conditioner is operating, and acquires the second information and the first information by extracting them from the acquired air conditioner information.

[0010] A third aspect of the present disclosure is the state determination device according to the first aspect, wherein the control unit predicts the first information under normal conditions using a trained model, and the trained model is generated using training data including the first information and the second information under normal conditions acquired in the air conditioner during operation to acquire the training data.

[0011] A fourth aspect of the present disclosure is the status determination device according to the third aspect, wherein the control unit acquires air conditioner information while the air conditioner is operating, and acquires the second information and the first information by extracting the second information and the first information specified by a user from the acquired air conditioner information, and predicts the first information under normal conditions using a trained model corresponding to the second information and the first information specified by the user.

[0012] A fifth aspect of the present disclosure is a state determination device according to the third aspect, wherein the learning data further includes third information, which is environmental information at the time the second information was acquired, and the third information is acquired by the air conditioner during operation to acquire the learning data, and the control unit predicts the first information under normal conditions based on the second information and the third information acquired during operation of the air conditioner.

[0013] A sixth aspect of the present disclosure is the status determination device according to the fifth aspect, wherein the second information and the third information acquired during operation of the air conditioner are the second information and the third information specified by a user, and the control unit predicts the first information under normal conditions using a trained model corresponding to the second information and the third information specified by the user.

[0014] A seventh aspect of the present disclosure is the state determination device according to the fifth aspect, wherein the environmental information includes at least one of a measured amount of indoor temperature, a measured amount of indoor humidity, a measured amount of outdoor temperature, and a measured amount of outdoor humidity.

[0015] An eighth aspect of the present disclosure is a state determination device according to the third aspect, wherein the learning data further includes fourth information, which is a target value when the second information is acquired, and the fourth information is acquired by the air conditioner during operation to acquire the learning data, and the control unit predicts the first information under normal conditions based on the second information and the fourth information acquired during operation of the air conditioner.

[0016] A ninth aspect of the present disclosure is the state determination device according to the eighth aspect, wherein the second information and the fourth information acquired during operation of the air conditioner are the second information and the fourth information specified by a user, and the control unit predicts the first information under normal conditions using a trained model corresponding to the second information and the fourth information specified by the user.

[0017] A tenth aspect of the present disclosure is the state determination device according to the eighth aspect, wherein the target value includes at least one of a target value of a condensing temperature, a target value of an evaporating temperature, and a target value of an indoor temperature.

[0018] An eleventh aspect of the present disclosure is a state determination device according to any one of the first to tenth aspects, wherein the control unit determines the state of the air conditioner by comparing the first information predicted for a normal state with the first information acquired during operation of the air conditioner.

[0019] A twelfth aspect of the present disclosure is the state determination device according to any one of the first to eleventh aspects, wherein the control amount includes at least one of a compressor rotation speed, an electronic expansion valve opening, an outdoor fan rotation speed, and an indoor fan rotation speed.

[0020] A thirteenth aspect of the present disclosure is a state determination device according to any one of the first to twelfth aspects, wherein the measured quantities include at least one of a measured quantity of indoor temperature, a measured quantity of indoor humidity, a measured quantity of outdoor temperature, a measured quantity of outdoor humidity, a measured quantity of intake temperature, a measured quantity of discharge temperature, and a measured quantity of subcooling outlet temperature.

[0021] A fourteenth aspect of the present disclosure is a state determination device according to any one of the first to thirteenth aspects, wherein the first information in a predicted normal state includes an index value derived based on the first information in a predicted normal state.

[0022] A fifteenth aspect of the present disclosure is a state determination method in which a control unit provided in a state determination device for an air conditioner including monitored equipment and non-monitored equipment executes a process to determine the state of the air conditioner based on first information which is a control quantity or a measurement quantity related to the monitored equipment and second information which is a control quantity related to the non-monitored equipment, wherein the control unit executes a process to predict the first information in a normal state based on the second information acquired while the air conditioner is operating, and determine the state of the air conditioner based on the first information acquired while the air conditioner is operating and the predicted first information in a normal state.

[0023] According to the fifteenth aspect of the present disclosure, it is possible to improve the accuracy of determining the state of an air conditioner.

[0024] A sixteenth aspect of the present disclosure is a determination program for causing a control unit provided in a state determination device for an air conditioner including monitored equipment and non-monitored equipment to execute a process for determining the state of the air conditioner based on first information, which is a control quantity or measurement quantity related to the monitored equipment, and second information, which is a control quantity related to the non-monitored equipment, wherein the control unit executes a process for predicting the first information in a normal state based on the second information acquired while the air conditioner is operating, and determining the state of the air conditioner based on the first information acquired while the air conditioner is operating and the predicted first information in a normal state.

[0025] According to the sixteenth aspect of the present disclosure, it is possible to improve the accuracy of determining the state of an air conditioner.

[0026] FIG. 1 is a diagram illustrating an example of a system configuration of an air conditioning system. FIG. 2A is a first diagram illustrating an example of a refrigerant circuit of an air conditioner. FIG. 2B is a second diagram illustrating an example of a refrigerant circuit of an air conditioner. FIG. 2C is a third diagram illustrating an example of a refrigerant circuit of an air conditioner. FIG. 3A is a first diagram illustrating an example of various information acquired from an air conditioner. FIG. 3B is a second diagram illustrating an example of various information acquired from an air conditioner. FIG. 4 is a diagram illustrating an example of a hardware configuration of a state determination device. FIG. 5 is a diagram illustrating an example of a functional configuration of the state determination device in a learning phase. FIG. 6A is a first diagram illustrating a specific example of learning data. FIG. 6B is a second diagram illustrating a specific example of learning data. FIG. 6C is a third diagram illustrating a specific example of learning data. FIG. 6D is a fourth diagram illustrating a specific example of learning data. FIG. 6E is a fifth diagram illustrating a specific example of learning data. FIG. 6F is a sixth diagram illustrating a specific example of learning data. FIG. 7 is a first diagram illustrating an example of a functional configuration of the state determination device in a state determination phase. Fig. 8 is a diagram showing a specific example of processing by the determination unit. Fig. 9 is a flowchart showing the flow of learning processing. Fig. 10 is a first flowchart showing the flow of state determination processing. Fig. 11 is a second diagram showing an example of the functional configuration of the state determination device in the state determination phase. Fig. 12 is a second flowchart showing the flow of state determination processing. Fig. 13 is a third diagram showing an example of the functional configuration of the state determination device in the state determination phase. Fig. 14 is a third flowchart showing the flow of state determination processing.

[0027] Hereinafter, each embodiment will be described with reference to the accompanying drawings. In this specification and drawings, components having substantially the same functional configuration are designated by the same reference numerals, and redundant description will be omitted.

[0028] [First embodiment] <System configuration of air conditioning system> The system configuration of an air conditioning system including a state determination device according to a first embodiment will be described. Fig. 1 is a diagram showing an example of the system configuration of an air conditioning system. As shown in Fig. 1, the air conditioning system 100 has an air conditioner 110 and a state determination device 120. In the air conditioning system 100, the state determination device 120 is connected to the air conditioner so as to be able to communicate with it.

[0029] The air conditioner 110 includes various devices. In the first embodiment, the various devices included in the air conditioner 110 can be broadly divided into: monitored devices 111 that are monitored when determining the state (e.g., abnormality, degradation) of the air conditioner 110; and devices other than the monitored devices that are monitored when determining the state (e.g., abnormality, degradation) of the air conditioner 110 (non-monitored devices 112).

[0030] The monitored device 111 has at least an operating element or a sensor (having at least one of an operating element or a sensor, or both). The operating element operates based on a control amount (a control amount related to the monitored device 111) instructed by the controller 113. The sensor notifies the controller 113 of a measured amount that measures the operation of the operating element or a measured amount that measures a change in the process accompanying the operation of the operating element (a measured amount related to the monitored device 111).

[0031] The non-monitored device 112 has at least an operating element or a sensor (has at least one of an operating element or a sensor, or both). The operating element operates based on a control amount (a control amount related to the non-monitored device 112) instructed by the controller 113. The sensor notifies the controller 113 of a measured amount that measures the operation of the operating element or a measured amount that measures a change in the process accompanying the operation of the operating element (a measured amount related to the non-monitored device 112).

[0032] A target value is set in the controller 113. The controller 113 calculates a control amount based on the set target value and issues an instruction to an operating element of the monitored device 111 or an operating element of the non-monitored device 112. The controller 113 acquires a measured amount notified from a sensor of the monitored device 111 or a sensor of the non-monitored device 112. The controller 113 performs feedback control or sequence control on the monitored device 111 and the non-monitored device 112.

[0033] During the learning phase, the status determination device 120 collects information acquired by the controller 113 during normal operation of the air conditioner. Note that the example in Fig. 1 shows a case where the collection target is the controller 113 of the air conditioner 110. However, during the learning phase, the collection target from which the status determination device 120 collects information is not limited to the controller 113 of the air conditioner 110, and may be an air conditioner controller other than the controller 113 of the air conditioner 110.

[0034] The information collected by the status determination device 120 from the air conditioners includes first information, second information, and fourth information. The first information refers to a controlled variable or a measured variable related to the monitored device 111. The second information refers to a controlled variable related to the non-monitored device 112. The fourth information refers to a target value when the second information is acquired. Note that the "target value when the second information is acquired" is not strictly limited to the target value at the time the second information is acquired, but simply refers to a target value corresponding to the acquired second information. Hereinafter, similar expressions will be interpreted similarly.

[0035] During the learning phase, the state determination device 120 collects, as third information, environmental information acquired by an environmental sensor 130 outside the air conditioning system 100 while the air conditioner 110 is operating normally. Alternatively, during the learning phase, the state determination device 120 collects, as third information, weather information acquired via a network 140 outside the air conditioning system 100 while the air conditioner 110 is operating normally.

[0036] In the learning phase, the state determination device 120 acquires learning data by collecting the first information to the fourth information. Specifically, the state determination device 120 acquires learning data in which: one or more of the multiple control variables included in the second information in the normal state are used as an explanatory variable; and one or more of the multiple control variables included in the first information in the normal state are used as a target variable.

[0037] Alternatively, the state determination device 120 acquires learning data in which: - any one or two or more of the multiple control quantities included in the second information under normal conditions is used as an explanatory variable; and - any one or two or more of the multiple measurement quantities included in the first information under normal conditions is used as a target variable.

[0038] Alternatively, the state determination device 120 acquires learning data in which: - any one or two or more of the multiple control quantities included in the second information in normal times is used as an explanatory variable; and - any one or two or more of the multiple control quantities included in the first information in normal times and any one or two or more of the multiple measurement quantities included in the first information in normal times are used as target variables.

[0039] Alternatively, the state determination device 120 acquires learning data in which: - any one or more control variables among the multiple control variables included in the second information under normal conditions and any one or more measurement variables among the multiple measurement variables included in the third information acquired when the second information under normal conditions is acquired are used as explanatory variables; and - any one or more control variables among the multiple control variables included in the first information under normal conditions are used as the objective variable. Note that "third information acquired when the second information under normal conditions is acquired" is not strictly limited to third information acquired at the timing when the second information under normal conditions is acquired, but simply refers to third information corresponding to the acquired second information under normal conditions. Hereinafter, similar expressions will be interpreted similarly.

[0040] Alternatively, the state determination device 120 acquires learning data in which: - any one or more of the multiple control variables included in the second information under normal conditions and any one or more target values ​​included in the fourth information acquired when the second information under normal conditions are used as explanatory variables; and - any one or more of the multiple control variables included in the first information under normal conditions are used as objective variables. Note that "fourth information acquired when the second information under normal conditions is acquired" is not strictly limited to fourth information acquired at the timing when the second information under normal conditions is acquired, but simply refers to fourth information corresponding to the acquired second information under normal conditions. Hereinafter, similar expressions will be interpreted similarly.

[0041] Alternatively, the state determination device 120 acquires learning data in which: - any one or more control variables among the multiple control variables included in the second information under normal conditions and any one or more measurement variables among the multiple measurement variables included in the third information acquired when the second information under normal conditions is acquired are used as explanatory variables; and - any one or more measurement variables among the multiple measurement variables included in the first information under normal conditions are used as target variables.

[0042] Alternatively, the state determination device 120 acquires learning data in which: - any one or two or more of the multiple control variables included in the second information under normal conditions and any one or two or more of the multiple target values ​​included in the fourth information acquired when the second information under normal conditions are acquired are used as explanatory variables; and - any one or two or more of the multiple measurement variables included in the first information under normal conditions are used as target variables.

[0043] Alternatively, the state determination device 120 acquires learning data in which: - any one or more of the multiple control quantities included in the second information under normal conditions and any one or more of the multiple measurement quantities included in the third information acquired when the second information under normal conditions are used as explanatory variables; and - any one or more of the multiple control quantities included in the first information under normal conditions and any one or more of the multiple measurement quantities included in the first information under normal conditions are used as objective variables.

[0044] Alternatively, the state determination device 120 acquires learning data in which: - any one or more of the multiple control quantities included in the second information under normal conditions and any one or more target values ​​included in the fourth information acquired when the second information under normal conditions are used as explanatory variables; and - any one or more of the multiple control quantities included in the first information under normal conditions and any one or more measurement quantities included in the first information under normal conditions are used as objective variables.

[0045] Alternatively, the state determination device 120 acquires learning data in which: - any one or two or more of the multiple control variables included in the second information under normal conditions, and any one or two or more of the multiple measurement variables included in the third information acquired when the second information under normal conditions was acquired, and any one or two or more of the multiple target values ​​included in the fourth information, are used as explanatory variables; and - any one or two or more of the multiple control variables included in the first information under normal conditions are used as the objective variable.

[0046] Alternatively, the state determination device 120 acquires learning data in which: - any one or two or more of the multiple control quantities included in the second information under normal conditions, and any one or two or more of the multiple measurement quantities included in the third information acquired when the second information under normal conditions was acquired, and any one or two or more of the multiple target values ​​included in the fourth information, are used as explanatory variables; and - any one or two or more of the multiple measurement quantities included in the first information under normal conditions are used as the objective variable.

[0047] Alternatively, the state determination device 120 acquires learning data in which: - any one or two or more of the multiple control quantities included in the second information under normal conditions, and any one or two or more of the multiple measurement quantities included in the third information acquired when the second information under normal conditions was acquired, and any one or two or more of the multiple target values ​​included in the fourth information, are used as explanatory variables; and - any one or two or more of the multiple control quantities included in the first information under normal conditions, and any one or two or more of the measurement quantities included in the first information under normal conditions are used as objective variables.

[0048] In the learning phase, the state determination device 120 performs a learning process on each learning model using each acquired learning data, and generates each trained model.

[0049] In the state determination phase, the state determination device 120 determines a trained model corresponding to: - information on a prediction target determined from among a plurality of control variables or a plurality of measurement variables included in the first information; and - information on an input target determined from among a plurality of control variables included in the second information (and the third information, or and the fourth information, or and the third and fourth information).

[0050] In the state determination phase, the state determination device 120 reads out the determined trained model and predicts information (predicted value) of the prediction target in normal times by inputting information (actual measured value) of the input target newly acquired during operation of the air conditioner 110. The state determination device 120 determines the state of the air conditioner 110 by comparing the information (actual measured value) of the prediction target newly acquired during operation of the air conditioner 110 with the predicted information (predicted value) of the prediction target in normal times.

[0051] In this way, the state determination device 120: - In the learning phase, when acquiring learning data including explanatory variables and target variables determined from the first to fourth information acquired during normal operation of the air conditioner, the second information, which is the "control amount" related to the equipment not to be monitored, is used as the explanatory variable. - In the state determination phase, when predicting information (predicted value) of the prediction target under normal conditions using the trained model, the state determination device 120 uses the second information, which is the "control amount" related to the equipment not to be monitored, as information to be input.

[0052] As a result, the state determination device 120 can solve problems that arise in the following cases.

[0053] For example, we will explain a case in which, in the learning phase, "measurements" related to devices that are not subject to monitoring are used as explanatory variables, and, in the status determination phase, "measurements" related to devices that are not subject to monitoring are used as information to be input.

[0054] In the above case, when an abnormality or the like occurs in an air conditioner, the measured quantity (actual measured value) of the monitored device, which is monitored when determining the abnormality or the like, changes from the measured quantity (actual measured value) under normal conditions. At that time, the measured quantity (actual measured value) of devices other than the monitored device, which is monitored when determining the abnormality or the like, may also change. In such a case, the information of the predicted target under normal conditions (predicted value), which is predicted by inputting the changed measured quantity (actual measured value) of the non-monitored device as input target information into the trained model, will also deviate from the measured quantity (actual measured value) of the monitored device under normal conditions. In other words, the prediction accuracy will decrease. As a result, for example, there is a possibility that an air conditioner may be erroneously determined to be normal even when an abnormality or the like has occurred.

[0055] Alternatively, in the above case, if an abnormality occurs in a sensor measuring a measurement quantity (actual measurement value) of a non-monitored device, the measurement quantity (actual measurement value) of the non-monitored device changes. As a result, the information (predicted value) of the predicted target under normal conditions, which is predicted by inputting the measurement quantity (actual measurement value) of the non-monitored device as input target information into the trained model, also deviates from the measurement quantity (actual measurement value) of the non-monitored device under normal conditions. In other words, the prediction accuracy decreases. As a result, even if an abnormality or the like different from the abnormality or the like determined based on the measurement quantity (actual measurement value) of the monitored device has occurred, it may be erroneously determined that an abnormality or the like determined based on the measurement quantity (actual measurement value) of the monitored device has occurred. Note that an abnormality or the like different from the abnormality or the like determined based on the measurement quantity (actual measurement value) of the monitored device here refers to an abnormality in the sensor measuring the measurement quantity (actual measurement value) of the non-monitored device.

[0056] In contrast to this, the status determination device 120 uses, as information to be input, a "control amount" for the non-monitored device 112 instead of a "measurement amount" for the non-monitored device 112. As described above, the "control amount" for the non-monitored device 112 is calculated based on a set target value, and does not change even if the measurement amount for the non-monitored device 112 changes.

[0057] Alternatively, the status determination device 120 uses the third information or the fourth information as information to be input in addition to the "control amount" related to the non-monitored device 112. As is clear from the above explanation, the third information or the fourth information does not change even if an abnormality or the like occurs in the air conditioner 110.

[0058] Therefore, the information (predicted value) of the target to be predicted under normal conditions, which is predicted by inputting the control amount (and the third information, or the fourth information, or the third and fourth information) related to the non-monitored device 112 as input target information into the trained model, will not deviate from the true value. The true value here refers to the measured amount (actual value) related to the monitored device under normal conditions. In other words, according to the state determination device 120, even if: - the measured amount (actual value) related to the non-monitored device changes due to the occurrence of an abnormality or the like in the air conditioner, or - an abnormality occurs in the sensor for measuring the measured amount (actual value) related to the non-monitored device, it is possible to avoid erroneous determinations such as those described above, and improve the accuracy of determination when determining the state of the air conditioner.

[0059] <Configuration Example of Refrigerant Circuit of Air Conditioner> Next, a configuration example of the refrigerant circuit provided in the air conditioner 110 will be described.

[0060] (1) Refrigerant Circuit Configuration Example 1 Fig. 2A is a first diagram showing an example of a refrigerant circuit of an air conditioner. Refrigerant circuit 200 shown in Fig. 2A includes a compressor 201, a condenser 202, an expansion valve 203, an evaporator 204, an outdoor unit fan 205, and an indoor unit fan 206. Refrigerant circuit 200 includes a temperature sensor 211 (compressor discharge pipe temperature sensor), a temperature sensor 212 (high-pressure side heat exchanger inlet gas pipe temperature sensor), and a pressure sensor 213 (high-pressure side pressure sensor). Refrigerant circuit 200 also includes a temperature sensor 214 (high-pressure side heat exchanger outlet liquid pipe temperature sensor) and a temperature sensor 215 (compressor suction pipe temperature sensor).

[0061] (2) Refrigerant Circuit Configuration Example 2 Fig. 2B is a second diagram showing an example of a refrigerant circuit of an air conditioner. Refrigerant circuit 220 shown in Fig. 2B is a refrigerant circuit for a chiller, and includes a compressor 221, a condenser 222, and an economizer circuit 223. The economizer circuit 223 includes an economizer heat exchanger 223_1 and an economizer heat exchanger expansion valve 223_2.

[0062] The refrigerant circuit 220 includes an expansion valve 224, an evaporator 225, an indoor heat exchanger 226, a pump 227, an outdoor unit fan 228, and an indoor unit fan 229. The refrigerant circuit 220 also includes a temperature sensor 231, a temperature sensor 232 (high-pressure side heat exchanger inlet gas pipe temperature sensor), and a temperature sensor 233 (high-pressure side heat exchanger outlet liquid pipe temperature sensor).

[0063] The refrigerant circuit 220 includes a temperature sensor 234, a temperature sensor 235 (economizer outlet (bypass side) temperature sensor), and a pressure sensor 236 (high-pressure side pressure sensor).

[0064] (3) Refrigerant Circuit Configuration Example 3 Figure 2C is a third diagram showing an example of a refrigerant circuit for an air conditioner. Refrigerant circuit 240 shown in Figure 2C is a refrigerant circuit for a multi-air conditioner for buildings, and includes a compressor 241, a four-way valve 242, a condenser 243, a condenser outlet expansion valve 244, and a supercooling circuit 245. Note that the supercooling circuit 245 includes a supercooling heat exchanger 245_1 and a supercooling heat exchanger expansion valve 245_2.

[0065] The refrigerant circuit 240 includes evaporator inlet expansion valves 246_1 to 246_n, evaporators 247_1 to 247_n, an accumulator 248, an outdoor unit fan 249, and an indoor unit fan 250.

[0066] The refrigerant circuit 240 includes a temperature sensor 251 (compressor discharge pipe temperature sensor), a temperature sensor 252 (high-pressure side heat exchanger inlet gas pipe temperature sensor), a temperature sensor 253 (high-pressure side heat exchanger outlet pipe temperature sensor), and a temperature sensor 254 .

[0067] The refrigerant circuit 260 includes a temperature sensor 255 (subcooling heat exchanger outlet gas pipe temperature sensor), a temperature sensor 256 (compressor suction pipe temperature sensor), and a pressure sensor 257 (high-pressure side pressure sensor).

[0068] <Various types of information acquired from air conditioners> Next, specific examples of the various types of information acquired by the state determination device 120 from the air conditioners 110 will be described. Fig. 3A is a first diagram showing an example of the various types of information acquired from the air conditioners, and is an example of the various types of information acquired from the air conditioner 110 having the refrigerant circuit 240 shown in Fig. 2C. As described above, the first information refers to the controlled variable or measured variable related to the equipment to be monitored. As shown in FIG. 3A , control quantities related to the equipment to be monitored include, for example, the control quantity of the condenser outlet expansion valve opening (an example of an electronic expansion valve opening), the control quantity of the evaporator inlet expansion valve opening (an example of an electronic expansion valve opening), and the control quantity of the subcooling expansion valve opening (an example of an electronic expansion valve opening). Measurement quantities related to the equipment to be monitored include, for example, the measurement quantity of the compressor suction pipe temperature (an example of an suction temperature), the measurement quantity of the compressor discharge pipe temperature (an example of a discharge temperature), the measurement quantity of the high-pressure side heat exchanger inlet gas pipe temperature, the measurement quantity of the high-pressure side heat exchanger outlet pipe temperature, and the measurement quantity of the subcooling heat exchanger outlet gas pipe temperature (an example of a subcooling outlet temperature).

[0069] As described above, the second information refers to control variables related to devices that are not monitored. As shown in Fig. 3A, the control variables related to devices that are not monitored include, for example, the control variable of the compressor rotation speed, the control variable of the outdoor unit fan step (an example of the outdoor fan rotation speed), and the control variable of the indoor unit fan step (an example of the indoor fan rotation speed).

[0070] 3B is a second diagram showing an example of various information acquired from the air conditioner. As described above, the third information refers to environmental information acquired by an environmental sensor 130 outside the air conditioning system 100, or weather information acquired via a network 140 outside the air conditioning system 100. As shown in FIG. 3B , the environmental information or weather information includes, for example, a measured amount of indoor temperature, a measured amount of indoor humidity, a measured amount of outdoor temperature, and a measured amount of outdoor humidity.

[0071] As described above, the fourth information refers to the target values ​​when the second information is acquired. As shown in Fig. 3B , the target values ​​when the second information is acquired include, for example, the target value of the condensing temperature, the target value of the evaporating temperature, and the target value of the indoor temperature (the set temperature of the indoor unit).

[0072] <Hardware Configuration of State Determination Device> Next, a description will be given of the hardware configuration of the state determination device 120. Fig. 4 is a diagram showing an example of the hardware configuration of the state determination device.

[0073] 4, the state determination device 120 includes a processor 401, a memory 402, an auxiliary storage device 403, an I / F (Interface) device 404, a communication device 405, and a drive device 406. The hardware components of the state determination device 120 are connected to each other via a bus 407.

[0074] The processor 401 has various computing devices such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), etc. The processor 401 reads various programs (for example, a state determination program) into the memory 402 and executes them.

[0075] The memory 402 has a main storage device such as a ROM (Read Only Memory), a RAM (Random Access Memory), etc. The processor 401 and the memory 402 form a so-called computer (also referred to as a "control unit 400"), and the processor 401 executes various programs read onto the memory 402, causing the computer to realize various functions.

[0076] The auxiliary storage device 403 stores various programs and various information used when the processor 401 executes the various programs.

[0077] The I / F device 404 connects the state determination device 120 with an operation device 411 for inputting user instructions and a display device 412 for displaying processing results to the user.

[0078] The communication device 405 is connected to an internal network (not shown) and performs communication processing with the air conditioner 110, the environmental sensor 130, etc. Alternatively, the communication device 405 is connected to an external network 140 and performs communication processing with an external server device (not shown), etc.

[0079] The drive device 406 is a device for setting a recording medium 413. The recording medium 413 here includes media that record information optically, electrically, or magnetically, such as a CD-ROM, a flexible disk, a magneto-optical disk, etc. The recording medium 413 may also include semiconductor memories that record information electrically, such as ROMs, flash memories, etc.

[0080] The various programs to be installed in the auxiliary storage device 403 are installed, for example, by setting the distributed recording medium 413 in the drive device 406 and reading the various programs recorded on the recording medium 413 by the drive device 406. Alternatively, the various programs to be installed in the auxiliary storage device 403 may be installed by being downloaded from the external network 140 via the communication device 405.

[0081] <Functional Configuration of State Determination Device (Learning Phase)> Next, a description will be given of functions that are realized in the state determination device 120 by executing the state determination program. First, a description will be given of functions in the learning phase.

[0082] 5 is a diagram showing an example of the functional configuration of the state determination device 120 in the learning phase. As shown in FIG. 5, in the learning phase, the state determination device 120 functions as an air conditioner information acquisition unit 510, a learning data acquisition unit 520, an external information acquisition unit 530, and a learning unit 540.

[0083] The air conditioner information acquisition unit 510 collects information (first information, second information, fourth information) acquired by the controller 113 during normal operation of the air conditioner 110. The air conditioner information acquisition unit 510 notifies the learning data acquisition unit 520 of the collected information (first information, second information, fourth information).

[0084] The external information acquisition unit 530 collects third information acquired when the first information, second information, and fourth information are acquired by the controller 113 during normal operation of the air conditioner 110. The external information acquisition unit 530 notifies the learning data acquisition unit 520 of the collected third information.

[0085] The learning data acquisition unit 520 determines combinations of explanatory variables and objective variables based on the first to fourth information, and acquires learning data. As described above, the learning data acquired by the learning data acquisition unit 520 includes various combinations of explanatory variables and objective variables, and the acquired learning data corresponds to the content of the state to be determined (what state is to be determined).

[0086] The learning data acquisition unit 520 stores the acquired multiple pieces of learning data in the learning data storage unit 550 .

[0087] The learning unit 540 has a learning model corresponding to each of the one or more acquired learning data, and a comparison and change unit. The learning unit 540 reads one piece of learning data from the one or more pieces of learning data stored in the learning data storage unit 550, and performs a learning process on the corresponding learning model to generate a trained model that predicts the first information in a normal state. The learning unit 540 generates a trained model corresponding to each of the one or more pieces of learning data, and stores the generated one or more trained models in the trained model storage unit 560.

[0088] The example of Fig. 5 shows how the learning unit 540 reads out learning data 551 from the learning data storage unit 550, sequentially inputs explanatory variables included in the read learning data 551 into the learning model 541, and the learning model 541 outputs output data. The example of Fig. 5 shows how the learning unit 540 sequentially inputs objective variables included in the read learning data 551 into the comparison and change unit 542. The example of Fig. 5 shows how the comparison and change unit 542 compares the input objective variables with the output data output from the learning model 541 and updates the model parameters of the learning model 541 according to the error.

[0089] In the above description, the learning data acquisition unit 520 is configured to store learning data, but the learning data acquisition unit 520 may also be configured to store the first information to the fourth information in the learning data storage unit 550. In this case, however, when reading the first information to the fourth information from the learning data storage unit 550, the learning data acquisition unit 520 acquires learning data by, for example, performing a filter process on the first information to the fourth information, and inputs the acquired learning data to the learning unit 540.

[0090] <Specific Examples of Learning Data> Next, a description will be given of specific examples of learning data acquired by the learning data acquisition unit 520. Fig. 6A is a first diagram showing specific examples of learning data.

[0091] As described above, the learning data acquisition unit 520 acquires learning data in the following manner: - one or more of the multiple control variables included in the second information in the normal state is used as an explanatory variable; - one or more of the multiple control variables included in the first information in the normal state is used as a response variable. In Figure 6A, learning data example 611 shows learning data in the following manner: - the devices to be monitored when determining the state of the air conditioner 110 are the evaporator inlet expansion valve and the subcooling expansion valve, and - the device other than the devices to be monitored when determining the state of the air conditioner 110 (device not to be monitored) is the compressor.

[0092] As described above, the learning data acquisition unit 520 acquires learning data in the following manner: - one or more of the multiple controlled variables included in the second information during normal operation is used as an explanatory variable; - one or more of the multiple measured variables included in the first information during normal operation is used as a response variable. In Figure 6A, learning data example 612 shows learning data in the following manner: - the monitored device monitored when determining the state of the air conditioner 110 is a subcooling heat exchanger outlet gas pipe temperature sensor; - the device other than the monitored device (non-monitored device) monitored when determining the state of the air conditioner 110 is a compressor; - the controlled variable of the compressor rotation speed is used as an explanatory variable; - the measured value of the subcooling heat exchanger outlet gas pipe temperature is used as a response variable.

[0093] As described above, the learning data acquisition unit 520 acquires learning data in the following manner: (1) any one or more of the multiple control variables included in the second information during normal operation is used as an explanatory variable; and (2) any one or more of the multiple control variables included in the first information during normal operation and any one or more of the multiple measurement variables included in the first information during normal operation are used as response variables. In Figure 6A, learning data example 613 shows learning data in the following manner: (1) the devices monitored when determining the state of the air conditioner 110 are the evaporator inlet expansion valve, the subcooling expansion valve, and the subcooling heat exchanger outlet gas pipe temperature sensor; and (2) the device other than the devices monitored when determining the state of the air conditioner 110 (device not monitored) is the compressor. (3) the control variable of the compressor rotation speed is used as an explanatory variable; and (4) the control variable of the evaporator inlet expansion valve opening, the control variable of the subcooling expansion valve opening, and the measurement variable of the subcooling heat exchanger outlet gas pipe temperature are used as response variables.

[0094] 6B is a second diagram showing a specific example of the learning data. As described above, the learning data acquiring unit 520 acquires learning data in which: any one or more of the plurality of control variables included in the second information under normal conditions and any one or more of the plurality of measurement variables included in the third information acquired when the second information under normal conditions are acquired are used as explanatory variables; and any one or more of the plurality of control variables included in the first information under normal conditions are used as a response variable. In Figure 6B, example learning data 621 shows learning data in the following case: - the equipment to be monitored when determining the state of the air conditioner 110 is the evaporator inlet expansion valve and the subcooling expansion valve, - the equipment other than the equipment to be monitored when determining the state of the air conditioner 110 (equipment not to be monitored) is the compressor and the environmental sensor, - the control amount of the compressor rotation speed, the measured amount of indoor temperature, and the measured amount of indoor humidity are the explanatory variables, - the control amount of the evaporator inlet expansion valve opening and the control amount of the subcooling expansion valve opening are the objective variables.

[0095] As described above, the learning data acquisition unit 520 acquires learning data in which: the explanatory variables are one or more of the multiple control variables included in the second information during normal operation and one or more of the multiple target values ​​included in the fourth information acquired when the second information during normal operation was acquired; and the objective variable is one or more of the multiple control variables included in the first information during normal operation. In Figure 6B, learning data example 622 shows learning data in which: the devices monitored when determining the state of the air conditioner 110 are the evaporator inlet expansion valve and the subcooling expansion valve, and the device other than the devices monitored when determining the state of the air conditioner 110 (the non-monitored device) is the compressor, the explanatory variables are the control variables of the compressor rotation speed, the target value of the condensing temperature, the target value of the evaporating temperature, and the target value of the indoor temperature are the control variables of the evaporator inlet expansion valve opening and the control variable of the subcooling expansion valve opening.

[0096] 6C is a third diagram showing a specific example of training data. As described above, the training data acquisition unit 520 acquires training data in which: one or more of the multiple controlled variables included in the second information during normal operation and one or more of the multiple measured variables included in the third information acquired when the second information during normal operation are used as explanatory variables; and one or more of the multiple measured variables included in the first information during normal operation are used as the response variable. In FIG. 6C, training data example 631 shows training data in which: the monitored device monitored when determining the state of the air conditioner 110 is a subcooling heat exchanger outlet gas pipe temperature sensor; and the devices other than the monitored device (non-monitored devices) monitored when determining the state of the air conditioner 110 are a compressor and an environmental sensor; the controlled variable of the compressor rotation speed, the measured value of the indoor temperature, and the measured value of the indoor humidity are used as explanatory variables; and the measured value of the subcooling heat exchanger outlet gas pipe temperature is used as the response variable.

[0097] As described above, the learning data acquisition unit 520 acquires learning data in which: (1) any one or more of the multiple controlled variables included in the second information during normal operation and any one or more of the multiple target values ​​included in the fourth information acquired when the second information during normal operation are used as explanatory variables; and (2) any one or more of the multiple measured variables included in the first information during normal operation are used as the objective variable. In Figure 6C, learning data example 632 shows learning data in which: (1) the monitored device monitored when determining the state of the air conditioner 110 is a subcooling heat exchanger outlet gas pipe temperature sensor; and (2) the device other than the monitored device (non-monitored device) monitored when determining the state of the air conditioner 110 is a compressor; (3) the controlled variable of the compressor rotation speed, the target value of the condensing temperature, the target value of the evaporating temperature, and the target value of the indoor temperature are used as explanatory variables; and (4) the measured value of the subcooling heat exchanger outlet gas pipe temperature is used as the objective variable.

[0098] 6D is a fourth diagram showing a specific example of the learning data. As described above, the learning data acquiring unit 520 acquires learning data in which: any one or more of the plurality of control variables included in the second information in the normal state and any one or more of the plurality of measurement variables included in the third information acquired when the second information in the normal state are acquired are used as explanatory variables; and any one or more of the plurality of control variables included in the first information in the normal state and any one or more of the plurality of measurement variables included in the first information in the normal state are used as response variables. In Figure 6D, example learning data 641 shows learning data in the following case: - the equipment to be monitored when determining the state of the air conditioner 110 is the evaporator inlet expansion valve, the subcooling expansion valve, and the subcooling heat exchanger outlet gas pipe temperature sensor; - the equipment other than the equipment to be monitored when determining the state of the air conditioner 110 (equipment not to be monitored) is the compressor and the environmental sensor; - the control amount of the compressor rotation speed, the measured amount of indoor temperature, and the measured amount of indoor humidity are the explanatory variables; - the control amount of the evaporator inlet expansion valve opening, the control amount of the subcooling expansion valve opening, and the measured amount of the subcooling heat exchanger outlet gas pipe temperature are the dependent variables.

[0099] As described above, the learning data acquisition unit 520 acquires learning data in which: - any one or more of the multiple control variables included in the second information under normal conditions and any one or more target values ​​included in the fourth information acquired when the second information under normal conditions are used as explanatory variables; and - any one or more of the multiple control variables included in the first information under normal conditions and any one or more measurement variables included in the first information under normal conditions are used as objective variables. In Figure 6D, example learning data 642 shows learning data in the following case: - the equipment to be monitored when determining the state of the air conditioner 110 is the evaporator inlet expansion valve, the subcooling expansion valve, and the subcooling heat exchanger outlet gas pipe temperature sensor, and - the equipment other than the equipment to be monitored when determining the state of the air conditioner 110 (equipment not to be monitored) is the compressor, - the control amount of the compressor rotation speed, the target value of the condensing temperature, the target value of the evaporating temperature, and the target value of the indoor temperature are the explanatory variables, and - the control amount of the evaporator inlet expansion valve opening, the control amount of the subcooling expansion valve opening, and the measured amount of the subcooling heat exchanger outlet gas pipe temperature are the dependent variables.

[0100] 6E is a fifth diagram showing a specific example of learning data. As described above, the learning data acquiring unit 520 acquires learning data in which: any one or more of the plurality of control variables included in the second information in the normal state, any one or more of the plurality of measurement variables included in the third information acquired when the second information in the normal state was acquired, and any one or more of the plurality of target values ​​included in the fourth information are used as explanatory variables; and any one or more of the plurality of control variables included in the first information in the normal state are used as objective variables. In Figure 6E, example learning data 651 shows learning data in the following case: - the equipment to be monitored when determining the state of the air conditioner 110 is the evaporator inlet expansion valve and the subcooling expansion valve, and - the equipment other than the equipment to be monitored when determining the state of the air conditioner 110 (equipment not to be monitored) is the compressor and the environmental sensor, - the control amount of the compressor rotation speed, the measured amount of the indoor temperature, the measured amount of the indoor humidity, the target value of the condensation temperature, the target value of the evaporation temperature, and the target value of the indoor temperature are the explanatory variables, and - the control amount of the evaporator inlet expansion valve opening and the control amount of the subcooling expansion valve opening are the objective variables.

[0101] As described above, the learning data acquisition unit 520 acquires learning data in which: - any one or two or more of the multiple control quantities included in the second information under normal conditions, any one or two or more of the multiple measurement quantities included in the third information acquired when the second information under normal conditions was acquired, and any one or two or more of the multiple target values ​​included in the fourth information are used as explanatory variables; and - any one or two or more of the multiple measurement quantities included in the first information under normal conditions are used as the objective variable. In Figure 6E, example learning data 652 shows learning data in the following case: - the monitored equipment to be monitored when determining the state of the air conditioner 110 is the subcooling heat exchanger outlet gas pipe temperature sensor, - the equipment other than the monitored equipment to be monitored when determining the state of the air conditioner 110 (equipment not to be monitored) is a compressor and an environmental sensor, - the control amount of the compressor rotation speed, the measured amount of indoor temperature, the measured amount of indoor humidity, the target value of condensation temperature, the target value of evaporation temperature, and the target value of indoor temperature are the explanatory variables, and - the measured amount of subcooling heat exchanger outlet gas pipe temperature is the dependent variable.

[0102] 6F is a sixth diagram showing a specific example of the learning data. As described above, the learning data acquiring unit 520 acquires learning data in the following manner: (1) any one or more of the plurality of control variables included in the second information in the normal state, (2) any one or more of the plurality of measurement variables included in the third information acquired when the second information in the normal state was acquired, and (3) any one or more of the plurality of target values ​​included in the fourth information are used as explanatory variables; and (4) any one or more of the plurality of control variables included in the first information in the normal state and (3) any one or more of the plurality of measurement variables included in the first information in the normal state are used as objective variables. In Figure 6F, example learning data 661 shows learning data in the following case: - the monitored equipment to be monitored when determining the state of the air conditioner 110 is the evaporator inlet expansion valve, the subcooling expansion valve, and the subcooling heat exchanger outlet gas pipe temperature sensor; - the equipment other than the monitored equipment to be monitored when determining the state of the air conditioner 110 (non-monitored equipment) is the compressor and the environmental sensor; - the control amount of the compressor rotation speed, the measured amount of indoor temperature, the measured amount of indoor humidity, the target value of condensation temperature, the target value of evaporation temperature, and the target value of indoor temperature are the explanatory variables; - the control amount of the evaporator inlet expansion valve opening, the control amount of the subcooling expansion valve opening, and the measured amount of the subcooling heat exchanger outlet gas pipe temperature are the dependent variables.

[0103] In the examples of Figures 6A to 6F, the objective variable is described as being the first information under normal conditions, but the objective variable is not limited to the first information under normal conditions and may be, for example, an index value derived based on the first information under normal conditions.

[0104] <Functional Configuration of State Determination Device (State Determination Phase)> Next, functions realized in the state determination device 120 in the state determination phase by executing the state determination program will be described.

[0105] 7 is a first diagram showing an example of the functional configuration of the state determination device 120 in the state determination phase. As shown in FIG. 7 , in the state determination phase, the state determination device 120 functions as an air conditioner information acquisition unit 710, an air conditioner information acquisition unit 711, a prediction unit 720, an external information acquisition unit 730, a determination unit 740, and an output unit 750.

[0106] The air conditioner information acquisition unit 710 acquires information determined as information to be input from the information (second information (and fourth information)) acquired by the controller 113 while the air conditioner 110 is operating. The air conditioner information acquisition unit 710 notifies the prediction unit 720 of the acquired information to be input. The air conditioner information acquisition unit 710 determines, as information to be input, information corresponding to the content of the state determination.

[0107] The external information acquisition unit 730 acquires third information from the environmental sensor 130 or from the network 140 while the air conditioner 110 is operating. If the acquired third information has been determined as information to be input, the external information acquisition unit 730 notifies the prediction unit 720. Note that the external information acquisition unit 730 determines whether or not to set the third information as information to be input, depending on the content of the state determination.

[0108] The prediction unit 720 reads out a trained model from the trained model storage unit 560, and inputs the information determined as the information to be input into the read trained model 721. The prediction unit 720 notifies the determination unit 740 of first information (predicted value) in normal conditions, which is the information to be predicted (predicted value) predicted by the trained model by inputting the information determined as the information to be input.

[0109] The prediction unit 720 reads out a trained model according to the content of the state to be determined. The example of FIG. 7 shows the state in which the prediction unit 720 reads out a trained model 721. When there are multiple content of state to be determined, multiple trained models corresponding to the multiple content of state to be determined are read out, and the state determination process is executed. The state determination process using multiple trained models may be executed in parallel or sequentially.

[0110] The air conditioner information acquisition unit 711 acquires information determined as information to be predicted from information (first information) acquired by the controller 113 while the air conditioner 110 is operating. The air conditioner information acquisition unit 711 notifies the determination unit 740 of the acquired first information (actual measured value), which is information (actual measured value) to be predicted. The air conditioner information acquisition unit 711 determines information corresponding to the content of the state determination as information to be predicted.

[0111] The determination unit 740 compares the first information (predicted value) in normal times, which is the information (predicted value) of the prediction target notified by the prediction unit 720, with the first information (actual value) which is the information (actual value) of the prediction target notified by the air conditioner information acquisition unit 711, to determine the state of the air conditioner 110. Specifically, the determination unit 740 calculates the degree of deviation between the first information (predicted value) in normal times, which is the information (predicted value) of the prediction target notified by the prediction unit 720, and the first information (actual value) which is the information (actual value) of the prediction target notified by the air conditioner information acquisition unit 711. The determination unit 740 determines the state of the air conditioner 110 based on the pattern of the calculated degree of deviation. The determination unit 740 notifies the output unit 750 of the state determination result.

[0112] The output unit 750 outputs the state determination result notified by the determination unit 740. The output method by the output unit 750 is arbitrary, and the output unit 750 may output the state determination result to the display device 412 connected to the state determination device 120, for example.

[0113] <Specific Example of Processing by Determination Unit> Next, a specific example of processing by the determination unit 740 of the state determination device 120 will be described. Fig. 8 is a diagram showing a specific example of processing by the determination unit. In Fig. 8, reference numeral 810 is a graph showing: a control amount (predicted value) of the evaporator inlet expansion valve opening, which is one of the first information (predicted values) in a normal state that is information (predicted values) to be predicted and that is notified from the prediction unit 720; and a control amount (actual measured value) of the evaporator inlet expansion valve opening, which is one of the first information (actual measured values) that is information (actual measured values) to be predicted and that is notified from the air conditioner information acquisition unit 711.

[0114] Also, in Figure 8, symbols 820 and 830 indicate how the degree of deviation between: - the first information (predicted value) under normal conditions, which is the information (predicted value) to be predicted and notified from the prediction unit 720; and - the first information (actual measured value), which is the information (actual measured value) to be predicted and notified from the air conditioner information acquisition unit 711, is calculated, and the state of the air conditioner 110 is determined based on the pattern of the calculated degree of deviation.

[0115] Specifically, in the example of Figure 8, the deviations are calculated for the refrigerant leakage determination index value (a value calculated based on measurement quantities), the control amount of the evaporator inlet expansion valve opening, the measured quantity of the subcooling heat exchanger outlet gas pipe temperature (subcooling outlet temperature), and the measured quantity of the compressor suction pipe temperature, and based on the pattern of the calculated deviations, the state determination result for the air conditioner 110 is output as "a state in which the refrigerant amount is decreasing (a state in which there is no gas)".

[0116] <Learning Process Flow> Next, a description will be given of the flow of the learning process performed by the state determination device 120. Fig. 9 is a flowchart showing the flow of the learning process.

[0117] In step S901, the state determination device 120 collects information (first information to fourth information) acquired by the controller while the air conditioner is operating normally.

[0118] In step S902, the state determination device 120 determines a combination of explanatory variables and response variables according to the state to be determined.

[0119] In step S903, the state determination device 120 acquires the learning data of the determined combination and stores it in the learning data storage unit 550.

[0120] In step S904, the state determination device 120 performs a learning process using the acquired learning data to generate a learned model.

[0121] In step S905, the state determination device 120 stores the generated trained model in the trained model storage unit 560.

[0122] In step S906, the state determination device 120 determines whether to generate a trained model based on another combination of training data. If it is determined in step S906 that a trained model based on another combination of training data is to be generated (YES in step S906), the process returns to step S902.

[0123] On the other hand, if it is determined in step S906 that a trained model based on another combination of training data is not to be generated (if NO in step S906), the training process ends.

[0124] <Flow of State Determination Process> Next, a description will be given of the flow of state determination process by the state determination device 120. Fig. 10 is a first flowchart showing the flow of state determination process.

[0125] In step S1001, the state determination device 120 determines information to be predicted from the first information according to the content of the state determination.

[0126] In step S1002, the state determination device 120 determines information to be input from the second information (and the third information, or the fourth information) according to the content of the state determination.

[0127] In step S1003, the state determination device 120 reads out a trained model corresponding to the determined information of the input target and the determined information of the prediction target.

[0128] In step S1004, the state determining device 120 acquires information (actual measurement values) of the determined input target while the air conditioner 110 is in operation.

[0129] In step S1005, the state determination device 120 acquires information (actual measurement values) of the determined prediction target while the air conditioner 110 is in operation.

[0130] In step S1006, the state determination device 120 inputs the acquired information (actual measured value) of the input target into the read trained model, thereby predicting information (predicted value) of the prediction target.

[0131] In step S1007, the state determination device 120 compares the information (actual measured value) of the prediction target acquired in step S1005 with the information (predicted value) of the prediction target predicted in step S1006, and determines the state of the air conditioner 110.

[0132] In step S1008, the state determination device 120 outputs the state determination result.

[0133] In step S1009, the state determination device 120 determines whether or not to continue the state determination process. If it is determined in step S1009 that the state determination process should be continued (YES in step S1009), the process returns to step S1004.

[0134] On the other hand, if it is determined in step S1009 that the state determination process should not be continued (NO in step S1009), the state determination process ends.

[0135] <Summary> As is clear from the above description, the state determination device 120 according to the first embodiment: is connected to air conditioners including monitored equipment and non-monitored equipment; determines the state of the air conditioner 110 based on first information (actual measured values), which is a controlled variable or measured variable related to the monitored equipment, and second information (actual measured values), which is a controlled variable related to the non-monitored equipment; specifically, predicts first information (predicted values) in normal conditions, which is information on the prediction target (predicted value), based on second information (actual measured values) acquired during operation of the air conditioner; determines the state of the air conditioner 110 based on first information (actual measured values), which is information on the prediction target (actual measured values) acquired during operation of the air conditioner, and first information (predicted values) in normal conditions, which is the predicted information on the prediction target (predicted value).

[0136] In this way, when predicting the first information under normal conditions from the second information, by using a "controlled variable" that is not affected by an abnormality as the second information (actual measured value), the first embodiment can improve the prediction accuracy when predicting the first information under normal conditions (predicted value). As a result, the first embodiment can improve the determination accuracy when determining the state of the air conditioner.

[0137] Second Embodiment In the above first embodiment, it has been described that the air conditioner information acquisition unit 710 acquires information determined as information to be input from among the information (second information (and fourth information)) acquired by the controller 113 while the air conditioner 110 is in operation. Also, in the above first embodiment, it has been described that the air conditioner information acquisition unit 711 acquires information determined as information to be predicted from among the information (first information) acquired by the controller 113 while the air conditioner 110 is in operation. In contrast, in the second embodiment, the air conditioner information acquisition unit acquires all air conditioner information acquired by the controller 113 while the air conditioner 110 is in operation, and extracts information determined as information to be input and information determined as information to be predicted. The second embodiment will be described below, focusing on the differences from the above first embodiment.

[0138] <Functional Configuration of State Determination Device (State Determination Phase)> First, the functional configuration of the state determination device according to the second embodiment in the state determination phase will be described. Fig. 11 is a second diagram showing an example of the functional configuration of the state determination device in the state determination phase. The difference from the functional configuration of the state determination device 120 according to the first embodiment in the state determination phase is that the state determination device 1100 according to the second embodiment has an air conditioner information acquisition unit 1110.

[0139] The air conditioner information acquisition unit 1110 acquires all air conditioner information acquired by the controller 113 while the air conditioner 110 is operating. The air conditioner information acquisition unit 1110 extracts information determined as information to be input from the second information (and fourth information) included in the acquired air conditioner information, and notifies the prediction unit 720. The air conditioner information acquisition unit 1110 extracts information determined as information to be predicted from the first information included in the acquired air conditioner information, and notifies the determination unit 740.

[0140] In this way, in the second embodiment, all of the air conditioner information acquired by the controller 113 while the air conditioner 110 is in operation is acquired once by the status determination device 1100, and information corresponding to the content of the status determination is extracted in the status determination device 1100. As a result, according to the second embodiment, it is not necessary for the controller 113 of the air conditioner 110 to execute processing to extract information corresponding to the content of the status determination.

[0141] <Flow of State Determination Process> Next, the flow of state determination process by state determination device 1100 according to the second embodiment will be described. Fig. 12 is a second flowchart showing the flow of state determination process. The difference from the flow of state determination process by state determination device 120 described in the first embodiment (Fig. 10) is steps S1201 to S1203.

[0142] In step S1201, the state determining device 1100 acquires all the air conditioner information acquired by the controller 113 while the air conditioner 110 is in operation.

[0143] In step S1202, the state determination device 1100 acquires the information determined as the information to be input. Note that the state determination device 1100 acquires the information to be input by extracting it from the second information (and fourth information) included in the acquired air conditioner information.

[0144] In step S1203, the state determination device 1100 acquires the information determined as the information to be predicted. Note that the state determination device 1100 acquires the information to be predicted by extracting it from the first information included in the acquired air conditioner information.

[0145] <Summary> As is clear from the above description, the state determination device 1100 according to the second embodiment: is connected to air conditioners including monitored and non-monitored equipment, and acquires all air conditioner information acquired by the air conditioners while the air conditioners are operating. From the acquired air conditioner information, it extracts first information (actual measured values) that is a controlled variable or measured variable related to the monitored equipment and second information (actual measured values) that is a controlled variable related to the non-monitored equipment, and determines the state of the air conditioners based on the first information (actual measured values) and the second information (actual measured values). Specifically, it predicts first information (predicted values) in normal conditions, which is information (predicted values) of a prediction target, based on the extracted second information (actual measured values). It determines the state of the air conditioners based on the first information (actual measured values) that is information (predicted values) of the extracted prediction target, and the first information (predicted values) in normal conditions, which is information (predicted values) of the predicted prediction target.

[0146] As a result, according to the second embodiment, it is possible to enjoy the same effects as the first embodiment, and further reduce the processing load on the air conditioner in the state determination process.

[0147] [Third Embodiment] In the first and second embodiments, the state determination process is described as being executed using determined input target information, determined prediction target information, and a trained model corresponding to the determined input target information and the determined prediction target information. In contrast, in the third embodiment, during execution of the state determination process, a user specifies a combination of input target information and prediction target information, and the state of the air conditioner is determined using a trained model corresponding to the specified combination. The third embodiment will be described below, focusing on the differences from the first and second embodiments.

[0148] <Functional Configuration of State Determination Device (State Determination Phase)> First, the functional configuration of the state determination device according to the third embodiment in the state determination phase will be described. Fig. 13 is a third diagram showing an example of the functional configuration of the state determination device in the state determination phase. The difference from the functional configuration of the state determination device 1100 according to the second embodiment in the state determination phase is that the state determination device 1300 according to the third embodiment has an air conditioner information acquisition unit 1310, a prediction unit 1320, and an external information acquisition unit 1330.

[0149] The air conditioner information acquisition unit 1310 receives from the user a designation of a combination of information to be input and information to be predicted. The air conditioner information acquisition unit 1310 acquires all air conditioner information acquired by the controller 113 while the air conditioner 110 is operating. The air conditioner information acquisition unit 1310 extracts information to be input designated by the user from the second information (and fourth information) included in the acquired air conditioner information, and notifies the prediction unit 720. The air conditioner information acquisition unit 1310 extracts information to be predicted designated by the user from the first information included in the acquired air conditioner information, and notifies the determination unit 740.

[0150] The external information acquisition unit 1330 acquires the third information from the environmental sensor 130 or from the network 140 while the air conditioner 110 is operating. The external information acquisition unit 730 extracts information to be input that is specified by the user from the acquired third information, and notifies the prediction unit 720 of the information.

[0151] The prediction unit 1320 receives, from the user, a designation of a combination of input target information and prediction target information. The prediction unit 1320 reads out a trained model corresponding to the combination designated by the user from the trained model storage unit 560, and inputs the input target information into the read trained model 721. The prediction unit 1320 notifies the determination unit 740 of first information (predicted value) in normal conditions, which is the information (predicted value) to be predicted by the trained model upon input of the input target information.

[0152] In this way, in the third embodiment, the state of the air conditioner is determined based on the combination specified by the user, which makes it possible to determine the state of the air conditioner from multiple perspectives.

[0153] <Flow of State Determination Process> Next, the flow of state determination process by the state determination device 1300 according to the third embodiment will be described. Fig. 14 is a third flowchart showing the flow of state determination process. The differences from the flow of state determination process by the state determination device 1100 described in the second embodiment (Fig. 12) are steps S1401 to S1403, S1404 to S1405, and S1406 to S1407.

[0154] In step S1401, the state determination device 1300 receives information of a prediction target designated by a user from the first information according to the content of the state determination.

[0155] In step S1402, the state determination device 1300 accepts information to be input designated by the user from the second information (and the third information, or the fourth information) according to the content of the state determination.

[0156] In step S1403, the state determination device 1300 reads out a trained model corresponding to the specified information of the input target and the specified information of the prediction target.

[0157] In step S1404, the status determination device 1300 acquires the information designated as the information to be input by extracting it from the second information (and fourth information) (and from the third information) included in the air conditioner information.

[0158] In step S1405, the state determining device 1300 obtains the information designated as the information to be predicted by extracting it from the first information included in the obtained air conditioner information.

[0159] In step S1406, the state determination device 1300 determines whether to continue the state determination process using the current trained model. If it is determined in step S1406 that the state determination process using the current trained model is to be continued (YES in step S1406), the process returns to step S1201.

[0160] On the other hand, if it is determined in step S1406 that the state determination process using the current trained model will not be continued (if NO in step S1406), proceed to step S1407.

[0161] In step S1407, the state determination device 1300 determines whether to continue the state determination process using another trained model. If it is determined in step S1407 that the state determination process using another trained model is to be continued (YES in step S1407), the process returns to step S1401.

[0162] On the other hand, if it is determined in step S1407 that the state determination process using other trained models will not be continued (if NO in step S1407), the state determination process is terminated.

[0163] <Summary> As is clear from the above description, the state determination device 1300 according to the third embodiment: is connected to air conditioners, including monitored and non-monitored devices, and acquires all air conditioner information acquired by the air conditioners while the air conditioners are operating. From the acquired air conditioner information, it extracts first information (actual measured values), which is a controlled variable or measured variable related to the monitored devices designated by the user, and second information (actual measured values), which is a controlled variable related to the non-monitored devices designated by the user. It then determines the state of the air conditioners based on the extracted first information (actual measured values) and second information (actual measured values). Specifically, it predicts first information (predicted values) in normal conditions, which is information related to a prediction target (predicted value), based on the extracted second information (actual measured values). It determines the state of the air conditioners based on the extracted first information (actual measured values), which is information related to a prediction target (predicted value), and the first information (predicted values) in normal conditions, which is information related to a predicted prediction target (predicted value).

[0164] As a result, according to the third embodiment, it is possible to obtain the same effects as in the second embodiment, and further to determine the state of the air conditioner from multiple perspectives.

[0165] [Fourth Embodiment] In the above first embodiment, it has been described that environmental information (measured amounts of indoor temperature, indoor humidity, outdoor temperature, outdoor humidity, etc.) is collected as third information from the environmental sensor 130 or the network 140. However, the third information may be environmental information acquired by a sensor in the air conditioning system 100. For example, a measured amount of outdoor temperature measured by a sensor in an outdoor unit or a measured amount of indoor temperature acquired by an indoor remote control may be acquired as third information from the controller 113. Note that when a measured amount measured by a sensor in the air conditioner 110 is acquired as environmental information from the controller 113, the environmental information may be acquired as first information.

[0166] Furthermore, in the first embodiment, no mention is made of the algorithm of the learning model, but the algorithm of the learning model is arbitrary, and for example, a neural network (NN) or the like may be used.

[0167] 1 shows the monitored devices 111 and the non-monitored devices 112, the distinction between the monitored devices 111 and the non-monitored devices 112 varies depending on the content of the status determination. For this reason, a single air conditioner 110 may be divided into multiple categories depending on the content of the status determination.

[0168] In the first embodiment, the learning process and the state determination process are performed by one state determination device 120. However, the processes may be performed by a plurality of state determination devices. Furthermore, the state determination device that performs the learning process and the state determination process may be separate devices.

[0169] Furthermore, in the above first embodiment, the state determination device 120 has been described as being provided separately from the air conditioner 110, but each function of the state determination device 120 may be realized, for example, in the controller 113 of the air conditioner 110. In other words, the state determination device 120 may be configured as an integral part of the air conditioner 110. Alternatively, even if the state determination device 120 is configured as a separate entity from the air conditioner 110, some of the functions of the state determination device 120 may be realized in the air conditioner 110.

[0170] Although not mentioned in the first embodiment, the state determination device 120 may be realized by, for example, a server device on the cloud, as long as it is capable of communicating with the air conditioner 110 .

[0171] Furthermore, in the first embodiment, examples of combinations of explanatory variables and response variables are shown using FIGS. 6A to 6D, but the combinations of explanatory variables and response variables are not limited to these.

[0172] Although the embodiments have been described above, it will be understood that various changes in form and details can be made without departing from the spirit and scope of the claims.

[0173] This application claims priority based on Japanese Patent Application No. 2024-091411, filed on June 5, 2024, the entire contents of which are incorporated herein by reference.

[0174] 100: Air conditioning system 110: Air conditioner 111: Equipment to be monitored 112: Equipment not to be monitored 113: Controller 120: State determination device 130: Environmental sensor 140: Network 510: Air conditioner information acquisition unit 520: Learning data acquisition unit 530: External information acquisition unit 540: Learning unit 710, 711: Air conditioner information acquisition unit 720: Prediction unit 721: Trained model 730: External information acquisition unit 740: Determination unit 750: Output unit 1100: State determination device 1110: Air conditioner information acquisition unit 1300: State determination device 1310: Air conditioner information acquisition unit 1320: Prediction unit 1330: External information acquisition unit

Claims

1. A state determination device (120, 1100, 1300) for an air conditioner (110) including monitored equipment and non-monitored equipment, comprising a control unit (400) that determines the state of the air conditioner based on first information, which is a controlled variable or a measured variable related to the monitored equipment, and second information, which is a controlled variable related to the non-monitored equipment, wherein the control unit (400) predicts the first information in a normal state based on the second information acquired while the air conditioner (110) is operating, and determines the state of the air conditioner based on the first information acquired while the air conditioner (110) is operating and the predicted first information in a normal state.

2. The state determination device (1100, 1300) according to claim 1, wherein the control unit (400) acquires air conditioner information while the air conditioner (110) is operating, and acquires the second information and the first information by extracting them from the acquired air conditioner information.

3. The state determination device (120, 1100, 1300) of claim 1, wherein the control unit (400) predicts the first information under normal conditions using a trained model, and the trained model is generated using training data including the first information and the second information under normal conditions acquired in the air conditioner (110) during operation to acquire the training data.

4. The state determination device (120, 1100, 1300) according to claim 3, wherein the learning data further includes third information, which is environmental information at the time the second information was acquired, and the third information is acquired by the state determination device (120, 1100, 1300) during operation to acquire the learning data, and the control unit (400) predicts the first information under normal conditions based on the second information and the third information acquired during operation of the air conditioner (110).

5. The state determination device (120, 1100, 1300) according to claim 4, wherein the environmental information includes at least one of a measured amount of indoor temperature, a measured amount of indoor humidity, a measured amount of outdoor temperature, and a measured amount of outdoor humidity.

6. The state determination device (120, 1100, 1300) according to claim 3, wherein the learning data further includes fourth information, which is a target value when the second information is acquired, and the fourth information is acquired by the air conditioner (110) during operation for acquiring the learning data, and the control unit (400) predicts the first information under normal conditions based on the second information and the fourth information acquired during operation of the air conditioner (110).

7. The state determination device (120, 1100, 1300) according to claim 6, wherein the target value includes at least one of a target value of a condensing temperature, a target value of an evaporating temperature, and a target value of an indoor temperature.

8. The state determination device (120, 1100, 1300) according to claim 1, wherein the control unit (400) determines the state of the air conditioner (110) by comparing the predicted first information in a normal state with the first information acquired while the air conditioner (110) is operating.

9. The state determination device (120, 1100, 1300) according to claim 1, wherein the control variable includes at least one of a compressor rotation speed, an electronic expansion valve opening, an outdoor fan rotation speed, and an indoor fan rotation speed.

10. The state determination device (120, 1100, 1300) according to claim 1, wherein the measured quantities include at least one of a measured quantity of indoor temperature, a measured quantity of indoor humidity, a measured quantity of outdoor temperature, a measured quantity of outdoor humidity, a measured quantity of intake temperature, a measured quantity of discharge temperature, and a measured quantity of subcooled outlet temperature.

11. A state determination device (120, 1100, 1300) according to claim 1, wherein the first information in the predicted normal state includes an index value derived based on the first information in the predicted normal state.

12. A state determination method in which a control unit (400) provided in a state determination device (120, 1100, 1300) of an air conditioner (110) including monitored equipment and non-monitored equipment executes a process to determine the state of the air conditioner based on first information which is a controlled variable or a measured variable related to the monitored equipment and second information which is a controlled variable related to the non-monitored equipment, wherein the control unit (400) executes a process to predict the first information in a normal state based on the second information acquired during operation of the air conditioner (110), and determine the state of the air conditioner based on the first information acquired during operation of the air conditioner (110) and the predicted first information in a normal state.

13. A status determination program for causing a control unit (400) provided in a status determination device (120, 1100, 1300) of an air conditioner (110) including equipment to be monitored and equipment not to be monitored to execute a process for determining the status of the air conditioner based on first information which is a controlled variable or a measured variable related to the equipment to be monitored and second information which is a controlled variable related to the equipment not to be monitored, the status determination program causing the control unit (400) to execute a process for predicting the first information in a normal state based on the second information acquired during operation of the air conditioner (110), and determining the status of the air conditioner based on the first information acquired during operation of the air conditioner (110) and the predicted first information in a normal state.

Images