Estimation method, program, and estimation system

Abstract

Provided is an estimation method in which first data which relates to the state of a device and which has a plurality of parameters including a specific parameter is acquired (S1). In the estimation method, second data is generated by changing the value of the specific parameter in the acquired first data (S2). In the estimation method, the acquired first data is input to an estimation model trained, using the abovementioned data as input, to output information indicating a change in the state of the device, and as result of the foregoing, first estimation data is output (S3). In the estimation method, second estimation data is output by inputting the second data to the estimation model (S4). In the estimation method, the impact of the specific parameter on the state of the device is estimated on the basis of the relationship between the first estimation data and the second estimation data (S5).

Description

Estimation method, program, and estimation system

[0001] The present disclosure relates to an estimation method for estimating an effect on a device.

[0002] For example, Patent Document 1 discloses an information processing system. This information processing system includes a diagnosis unit and a notification unit. The diagnosis unit diagnoses a fault of the refrigerator using a learning model that is trained to output a fault diagnosis result when information indicating an operating state of the refrigerator is input. The notification unit causes the refrigerator or a terminal device to output information based on the fault diagnosis result by the diagnosis unit.

[0003] Japanese Patent Application Laid-Open No. 2021-184132

[0004] The present disclosure provides an estimation method and the like that makes it easy to estimate the influence of parameters related to the state of equipment on the equipment.

[0005] An estimation method according to one aspect of the present disclosure acquires first data relating to the state of an equipment, the first data having a plurality of parameters including a specific parameter. The estimation method generates second data by changing the value of the specific parameter in the acquired first data. The estimation method outputs first estimated data by inputting the acquired first data into an estimation model that is trained to use the data as input and output information indicating changes in the state of the equipment. The estimation method outputs second estimated data by inputting the generated second data into the estimation model. The estimation method estimates the effect of the specific parameter on the state of the equipment based on the relationship between the first estimated data and the second estimated data.

[0006] A program according to one aspect of the present disclosure causes one or more processors to execute the estimation method.

[0007] An estimation system according to one aspect of the present disclosure includes an acquisition unit, a generation unit, a first estimation unit, a second estimation unit, and a third estimation unit. The acquisition unit acquires first data, which is data related to a state of an appliance and has a plurality of parameters including a specific parameter. The generation unit generates second data by changing the value of the specific parameter in the first data acquired by the acquisition unit. The first estimation unit outputs first estimated data by inputting the first data acquired by the acquisition unit into an estimation model that is trained to use the data as input and output information indicating a change in the state of the appliance. The second estimation unit outputs second estimated data by inputting the second data generated by the generation unit into the estimation model. The third estimation unit estimates an effect of the specific parameter on the state of the appliance based on a relationship between the first estimated data output by the first estimation unit and the second estimated data generated by the second estimation unit.

[0008] The estimation method and the like of the present disclosure have the advantage of making it easier to estimate the influence of parameters related to the state of the equipment on the equipment.

[0009] FIG. 1 is an explanatory diagram of a problem with a diagnostic method for diagnosing the state of a device. FIG. 2 is an explanatory diagram of another problem with a diagnostic method for diagnosing the state of a device. FIG. 3 is a block diagram showing an overall configuration including an estimation system according to an embodiment. FIG. 4 is a flowchart showing an example of operation of the estimation system according to an embodiment. FIG. 5 is a diagram schematically showing operation of the estimation system according to an embodiment. FIG. 6 is a diagram schematically showing a specific example of operation of the estimation system according to an embodiment.

[0010] [1. Findings that Form the Basis of the Present Disclosure] First, the inventor's viewpoint will be explained below.

[0011] In recent years, research has been conducted into diagnostic methods for diagnosing whether or not a device, such as a refrigerator, has malfunctioned when some abnormality occurs in the device. In particular, it is desirable for such diagnostic methods to not only determine whether or not the device has malfunctioned, but also to diagnose what caused the malfunction if the device has malfunctioned, or what caused the abnormality if the device has not malfunctioned.

[0012] Incidentally, for an appliance such as a refrigerator, multiple parameters related to the appliance state can affect the appliance state (e.g., the temperature inside the refrigerator). Here, the multiple parameters can include parameters related to user behavior with respect to the appliance, such as the number of times the refrigerator door is opened and closed. The multiple parameters can also include physical quantities representing the appliance state, such as the temperature of the refrigerator's refrigerator compartment, freezer compartment, or vegetable compartment. The multiple parameters can also include parameters related to appliance control, such as the frequency of the refrigerator's compressor. The multiple parameters can also include parameters representing the ambient conditions of the appliance, such as the room temperature where the refrigerator is installed.

[0013] FIG. 1 is an explanatory diagram of the problems with a diagnostic method for diagnosing the state of a device. The example shown in FIG. 1 illustrates a case where an abnormality occurs, such as an increase in the internal temperature after a user opens and closes a refrigerator door. Because multiple parameters can affect the state of a device as described above, it is difficult to identify whether the opening and closing of the refrigerator door is the cause of the increase in internal temperature. For example, the internal temperature may have increased because the room temperature in the location where the refrigerator is installed is relatively high. Furthermore, the internal temperature may have increased because the refrigerator's cooling performance was reduced due to excessive food or other items being packed into the refrigerator. Furthermore, the internal temperature may have increased because the refrigerator door was opened when cooling control was not being performed, such as during defrost control. Furthermore, if the refrigerator door had been opened and then immediately closed, the internal temperature may not have increased in the first place.

[0014] Here, one possible method for estimating which of multiple parameters affects equipment is to extract and verify only samples from a data set that collects samples having information on feature quantities and state changes made up of multiple parameters, where only the specific parameter whose effect is to be estimated differs and all of the other one or more parameters match. However, this method has the following problems.

[0015] FIG. 2 is an explanatory diagram of another problem with the method for diagnosing the state of a device. The data on the left side of FIG. 2 and the data on the right side represent an example of samples in which only the specific parameter, the number of times the refrigerator door is opened and closed, is different, and all of the other one or more parameters are the same. Here, there is a problem in that it is often difficult to find such samples in an actual refrigerator usage environment. Even if such samples are found, there is also a problem in that it is extremely difficult to prepare a sufficient number of samples to identify that the opening and closing of the refrigerator door is a factor affecting the refrigerator.

[0016] Furthermore, one possible method for diagnosing which of multiple parameters is a factor affecting an equipment is to construct an estimation model that is trained to input data containing multiple parameters, including a specific parameter, and output information indicating the effect that the specific parameter has on the state of the equipment. However, this method has a problem in that it is difficult to perform training because there is no correct answer data indicating the degree of effect of the specific parameter on changes in the state of the equipment.

[0017] In view of the above, the inventors have come up with the present disclosure.

[0018] Hereinafter, the embodiments will be described in detail with reference to the drawings. Note that the embodiments described below are all comprehensive or specific examples. The numerical values, shapes, materials, components, component placement and connection forms, steps, step order, etc. shown in the following embodiments are merely examples and are not intended to limit the present disclosure. Furthermore, among the components in the following embodiments, components not recited in independent claims will be described as optional components.

[0019] It should be noted that the drawings are schematic diagrams and are not necessarily strict illustrations. In addition, in the drawings, substantially the same components are denoted by the same reference numerals, and overlapping descriptions may be omitted or simplified.

[0020] [2. Configuration] The overall configuration including the estimation system 1 according to the embodiment will be described below with reference to Fig. 3. Fig. 3 is a block diagram showing the overall configuration including the estimation system 1 according to the embodiment. In addition to the estimation system 1, Fig. 3 also shows a device 2. The estimation system 1 is a system for diagnosing the state of the device 2.

[0021] The device 2 is, for example, a device installed in a facility where a user who uses the estimation system 1 resides. The device 2 is, for example, a home appliance. More specifically, the device 2 is, for example, a device that can be connected to a network N1 such as the Internet, and is a so-called IoT (Internet of Things) device. Specifically, the device 2 is, for example, a refrigerator, an air conditioner, a washing machine, a robot vacuum cleaner, a rice cooker, or a microwave oven. In the embodiment, the device 2 is a refrigerator, which is a device that operates continuously, in other words, a device that basically operates all the time. Note that the device 2 may not operate continuously. Note that the device 2 does not have to be an IoT device if it is equipped with the estimation system 1, as will be described later.

[0022] The estimation system 1 is a computer including a processor (e.g., a microprocessor) and a memory, etc. The memory is a read-only memory (ROM) and a random access memory (RAM), etc., and can store programs executed by the processor.

[0023] For example, the estimation system 1 may be a computer (device) in a single housing, or may be a system made up of multiple computers. For example, the estimation system 1 may be a server. Note that the components included in the estimation system 1 may be located in a single server, or may be distributed across multiple servers. Furthermore, for example, the estimation system 1 may be mounted on a device 2.

[0024] The estimation system 1 includes an acquisition unit 11, a generation unit 12, a first estimation unit 13, a second estimation unit 14, and a third estimation unit 15. The acquisition unit 11, the generation unit 12, the first estimation unit 13, the second estimation unit 14, and the third estimation unit 15 are each realized by a processor or the like that executes a program stored in a memory.

[0025] The acquisition unit 11 acquires first data. Here, the first data is data related to the state of the device 2 and includes a plurality of parameters including a specific parameter. In the embodiment, the first data is data related to the state of the refrigerator and includes parameters such as the room temperature of the location where the refrigerator is installed, the refrigerator's internal temperatures (such as the temperature of the refrigerator compartment, the temperature of the vegetable compartment, and the temperature of the freezer compartment), the compressor frequency, the number of times the refrigerator door is opened and closed, the open / close state of the damper, and the operating state of the defroster. These parameters are measured by the device 2 at regular intervals (for example, every 5 minutes). Note that the number of times the refrigerator door is opened and closed is the number of times it is opened and closed in a regular interval.

[0026] In the embodiment, the specific parameter and one or more parameters other than the specific parameter among the plurality of parameters included in the first data are not collinear. To confirm this, for example, a method of confirming that the absolute value of the correlation coefficient between the specific parameter and one or more parameters is equal to or less than a predetermined value can be considered.

[0027] In the embodiment, the specific parameter is a value related to an action of a user using the estimation system 1 with respect to the appliance 2. More specifically, in the embodiment, the specific parameter is a feature based on at least one of the number of times the refrigerator door is opened and closed and the duration of the door opening and closing. In the embodiment, only the number of times the refrigerator door is opened and closed is used as the feature. In the embodiment, the one or more parameters include at least one of a physical quantity representing the state of the appliance 2, a value related to control of the appliance 2, and a physical quantity representing a surrounding state of the appliance 2. More specifically, in the embodiment, the one or more parameters include an internal temperature of the refrigerator (such as the temperature of the refrigerator compartment, the temperature of the vegetable compartment, and the temperature of the freezer compartment) as a physical quantity representing the state of the appliance 2, and a compressor frequency as a value related to control of the appliance 2. In the embodiment, the one or more parameters include a room temperature in a location where the refrigerator is installed as a physical quantity representing a surrounding state of the appliance 2.

[0028] In the embodiment, the acquisition unit 11 acquires first data from the original data. In other words, the acquisition unit 11 extracts feature quantities from the original data to be used as input data for an estimation model 16 (described later). Here, the original data is data measured at regular intervals in the device 2 (hereinafter referred to as "measurement data"), and is a collection of a large amount of measurement data accumulated before the diagnosis by the estimation system 1 is performed. The estimation system 1 can receive the original data from a cloud where the measurement data of the device 2 is accumulated, for example, by wireless or wired communication via the network N1 using a communication interface.

[0029] The acquisition unit 11 acquires, as the first data, measurement data at a time specified by a user of the estimation system 1, or measurement data immediately prior to that time, from the original data, for example. Alternatively, the acquisition unit 11 may acquire, as the first data, the latest measurement data at the time when the diagnosis by the estimation system 1 is performed, from the original data, for example.

[0030] The generator 12 generates second data by changing the value of a specific parameter in the first data acquired by the acquirer 11. The second data is data in which only the specific parameter has been changed from the value of the specific parameter in the first data, and one or more parameters other than the specific parameter all match one or more parameters in the first data. In other words, the second data is pseudo data generated from the first data.

[0031] In the embodiment, the generation unit 12 generates the second data by changing the number of times the refrigerator door is opened and closed, which is a specific parameter in the first data acquired by the acquisition unit 11, to "0." Note that the value of the specific parameter in the second data does not have to be "0" as long as it is different from the value of the specific parameter in the first data.

[0032] The first estimation unit 13 outputs first estimation data by inputting the first data acquired by the acquisition unit 11 to the estimation model 16. Here, the estimation model 16 is a machine learning model that receives as input data having the above-described plurality of parameters (in other words, measurement data) and is trained to output information (predicted values) indicating changes in the state of the device 2. The estimation model 16 is configured using, for example, LightGBM (registered trademark), a neural network, or the like.

[0033] The estimation model 16 is constructed by supervised learning using a training dataset consisting of a large number of samples. Here, the training dataset includes measurement data as input data and changes in the state of the device 2 as correct answer data. In the embodiment, the measurement data, which is the input data, is data measured at regular intervals in the refrigerator. Also, in the embodiment, the changes in the state of the device 2, which is the correct answer data, are changes in the temperature inside the refrigerator. More specifically, the change in the temperature inside the refrigerator is the difference between the temperature inside the refrigerator (e.g., the temperature of the refrigerator compartment) in the measurement data and the temperature inside the refrigerator in the measurement data in the next cycle.

[0034] Therefore, in the embodiment, estimation model 16 is trained to use data measured in the refrigerator as input data and output information indicating a change in the refrigerator's internal temperature from the time the data was measured until a predetermined time (here, corresponding to a certain period) has elapsed. Therefore, in the embodiment, the first estimated data output by first estimator 13 indicates a change in the refrigerator's internal temperature from the time the first data was measured until a predetermined time has elapsed.

[0035] The second estimation unit 14 outputs the second estimated data by inputting the second data generated by the generation unit 12 to the estimation model 16. Here, the estimation model 16 used by the first estimation unit 13 and the estimation model 16 used by the second estimation unit 14 are the same model. Therefore, in the embodiment, the second estimated data output by the second estimation unit 14 indicates a change in the temperature inside the refrigerator from the time the second data is measured until a predetermined time has elapsed. Note that the second data is pseudo data that is not actually measured, and therefore the second estimated data is data that is obtained assuming that the second data is actually measured. Here, the time when the second data is measured may be interpreted as the time when the first data is measured.

[0036] The third estimation unit 15 estimates the effect of a specific parameter on the state of the equipment 2 based on the relationship between the first estimation data and the second estimation data obtained by comparing the first estimation data output by the first estimation unit 13 with the second estimation data output by the second estimation unit 14.

[0037] In the embodiment, the third estimation unit 15 estimates the above-mentioned influence of a difference between the value of a specific parameter in the first data and the value of a specific parameter in the second data based on the difference between the first estimation data and the second estimation data as the relationship between the first estimation data and the second estimation data. More specifically, in the embodiment, the third estimation unit 15 calculates the difference between the first estimation data and the second estimation data, i.e., the difference in the change in the refrigerator's internal temperature between data sets that differ in the specific parameter (here, the number of times the refrigerator door is opened and closed). The third estimation unit 15 also treats the calculated difference as the difference between the value of the specific parameter in the first data and the value of the specific parameter in the second data, i.e., the influence on the refrigerator's temperature change due to the difference in the number of times the refrigerator door is opened and closed between the original data and the pseudo data. As a result, the third estimation unit 15 estimates that the difference in the number of times the refrigerator door is opened and closed has affected the change in the refrigerator's internal temperature by the amount equal to the difference between the first estimation data and the second estimation data.

[0038] [3. Operation] The operation of the estimation system 1 according to the embodiment (i.e., the estimation method) will be described below with reference to Figs. 4 and 5. Fig. 4 is a flowchart showing an example of the operation of the estimation system 1 according to the embodiment. Fig. 5 is a diagram schematically showing the operation of the estimation system 1 according to the embodiment. Fig. 4(a) shows a series of operations of the estimation system 1, and Fig. 4(b) shows a specific example of a flow for diagnosing the state of a device by the estimation system 1.

[0039] 4A and 5, the acquisition unit 11 of the estimation system 1 acquires first data (S1). In the embodiment, as already described, the acquisition unit 11 acquires the first data from the original data (in other words, the acquisition unit 11 extracts, from the original data, feature quantities to be used as input data for the estimation model 16).

[0040] Next, as shown in Figures 4(a) and 5, the generator 12 of the estimation system 1 generates second data by changing the value of the specific parameter in the first data acquired by the acquirer 11 (S2). That is, the generator 12 generates the second data by changing only the value of the specific parameter whose influence is to be estimated in the first data to a pseudo specific parameter to be compared, while leaving the values ​​of the remaining one or more parameters unchanged. In the embodiment, as already described, the second data is generated by changing the value of the specific parameter in the first data to "0".

[0041] Next, as shown in Figures 4(a) and 5, the first estimation unit 13 of the estimation system 1 inputs the first data acquired by the acquisition unit 11 to the estimation model 16, thereby outputting first estimation data (S3). Also, as shown in Figures 4(a) and 5, the second estimation unit 14 of the estimation system 1 inputs the second data generated by the generation unit 12 to the estimation model 16, thereby outputting second estimation data (S4). Note that steps S3 and S4 may be executed in the reverse order, or may be executed in parallel.

[0042] Next, as shown in Figures 4A and 5, the third estimator 15 of the estimation system 1 estimates the influence of the specific parameter on the state of the device 2 based on the relationship between the first estimation data output by the first estimator 13 and the second estimation data output by the second estimator 14 (S5). Specifically, as shown in Figure 4B, in step S5, the third estimator 15 first calculates the difference between the value of the specific parameter in the first estimation data and the value of the specific parameter in the second estimation data (S51). Then, the third estimator 15 estimates the influence of the specific parameter on the state of the device 2 based on the calculated difference (S52).

[0043] Fig. 6 is a diagram schematically illustrating a specific example of the operation of the estimation system 1 according to the embodiment. The example illustrated in Fig. 6 is an example in which the device 2 is a refrigerator. In the example illustrated in Fig. 6, the specific parameter in the first data is the number of times the refrigerator door is opened and closed, and the one or more parameters in the first data include the room temperature in the location where the refrigerator is installed, the temperature of the refrigerator compartment, the temperature of the freezer compartment, the temperature of the vegetable compartment, and the frequency of the compressor.

[0044] In the example shown in FIG. 6 , the generation unit 12 of the estimation system 1 generates the second data by changing the number of times the refrigerator door has been opened and closed in the first data from "10 times" to "0 times." The first estimation unit 13 of the estimation system 1 inputs the first data into the estimation model 16, thereby outputting first estimation data indicating that the refrigerator's internal temperature (here, the temperature of the refrigerator compartment) will increase by 3 degrees Celsius after five minutes. The second estimation unit 14 of the estimation system 1 inputs the second data into the estimation model 16, thereby outputting second estimation data indicating that the refrigerator's internal temperature will increase by 1 degree Celsius after five minutes. The third estimation unit 15 of the estimation system 1 estimates that the refrigerator's internal temperature will increase by 2 degrees Celsius due to the opening and closing of the refrigerator door, based on the difference between the first estimation data and the second estimation data.

[0045] [4. Effects, etc.] As described above, in the estimation system 1 (estimation method) according to the embodiment, the first data and the second data, which differ only in the value of a specific parameter, are input to the same estimation model 16, and therefore it is possible to estimate the influence of only the specific parameter on changes in the state of the device 2. Therefore, the estimation system 1 according to the embodiment has the advantage of easily estimating which of multiple parameters is affecting the device 2.

[0046] Furthermore, in the estimation system 1 according to the embodiment, the second data is generated by changing a specific parameter in the acquired first data. In other words, the estimation system 1 according to the embodiment generates pseudo measurement data in which only the specific parameter is changed from the measurement data, which has the advantage that only the specific parameter described in [1. Knowledge forming the basis of the present disclosure] differs, and there is no need for samples in which one or more parameters other than the specific parameter are all the same.

[0047] Furthermore, in the estimation system 1 according to the embodiment, it is sufficient to prepare an estimation model 16 that is trained to receive input data having a plurality of parameters (in other words, measurement data) and output information indicating a change in the state of the device 2. Therefore, compared to constructing the estimation model described in [1. Knowledge Forming the Basis of the Present Disclosure], the estimation system 1 according to the embodiment has the advantage that it is easier to prepare a large number of training datasets required for training the estimation model 16 and to construct the estimation model 16. This is because, in the estimation system 1 according to the embodiment, when training the estimation model 16, it is sufficient to prepare a training dataset that includes measurement data as input data and changes in the state of the device 2 as ground truth data, and it is not necessary to prepare a training dataset that includes ground truth data indicating the degree of influence of a specific parameter on changes in the state of the device 2.

[0048] [5. Other Embodiments] As described above, the embodiments have been described as examples of the technology disclosed in this application. However, the technology in this disclosure is not limited to these, and can be applied to embodiments in which modifications, substitutions, additions, omissions, etc. are made as appropriate. Furthermore, it is also possible to combine the components described in the above embodiments to create new embodiments.

[0049] In the above embodiment, when the device 2 is a refrigerator, the specific parameter is the number of times the refrigerator door is opened and closed, but this is not limited to this. For example, the specific parameter may be the amount of contents in the refrigerator (total weight of food and other items stored in the refrigerator), the compressor frequency, whether a damper is open or closed, whether defrosting is performed, etc. Furthermore, the specific parameter is not limited to one type, and may be multiple types. Furthermore, the specific parameter may be changed as appropriate by a user of the estimation system 1.

[0050] For example, if the specific parameter is the internal volume of the refrigerator, the third estimation unit 15 can estimate the influence of the internal volume on a change in the internal temperature. Also, if the specific parameter is at least one of the compressor frequency, the opening / closing of the damper, and whether or not defrosting is performed, the third estimation unit 15 can estimate the influence of these specific parameters on a change in the internal temperature.

[0051] The order of the processes described in the above embodiments is merely an example. The order of multiple processes may be changed, or multiple processes may be executed in parallel. Furthermore, a process executed by a specific processing unit may be executed by another processing unit. Furthermore, part of the digital signal processing described in the above embodiments may be realized by analog signal processing.

[0052] For example, the present disclosure can be realized as a program for causing a computer (processor) to execute steps included in the estimation method. Furthermore, the present disclosure can be realized as a non-transitory computer-readable recording medium, such as a CD-ROM, on which the program is recorded.

[0053] For example, when the present disclosure is realized as a program (software), each step is performed by running the program using hardware resources such as a computer's CPU, memory, input / output circuitry, etc. In other words, each step is performed by the CPU acquiring data from memory or input / output circuitry, etc., performing calculations on the data, and outputting the calculation results to memory or input / output circuitry, etc.

[0054] In the above embodiment, each component included in the estimation system 1 may be configured with dedicated hardware or may be realized by executing a software program suitable for each component. Each component may be realized by a program execution unit such as a CPU or processor reading and executing a software program recorded on a recording medium such as a hard disk or semiconductor memory.

[0055] Some or all of the functions of the estimation system 1 according to the above embodiment are typically realized as an LSI, which is an integrated circuit. These may be individually integrated into single chips, or may be integrated into a single chip that includes some or all of the functions. Furthermore, the integrated circuit is not limited to an LSI, and may be realized using a dedicated circuit or a general-purpose processor. It is also possible to use an FPGA (Field Programmable Gate Array) that can be programmed after LSI manufacturing, or a reconfigurable processor that can reconfigure the connections and settings of circuit cells within an LSI.

[0056] Furthermore, if an integrated circuit technology that can replace LSI emerges due to advances in semiconductor technology or other derived technologies, it is natural that each component included in the estimation system 1 may be integrated into an integrated circuit using that technology.

[0057] In addition, this disclosure also includes forms obtained by making various modifications to the embodiments that a person skilled in the art would think of, and forms realized by arbitrarily combining the components and functions in each embodiment within the scope that does not deviate from the intent of this disclosure.

[0058] (Summary) As described above, the estimation method according to the first aspect acquires first data relating to the state of the device 2 and including a plurality of parameters including a specific parameter (S1). In this estimation method, second data is generated by changing the value of the specific parameter in the acquired first data (S2). In this estimation method, the acquired first data is input to an estimation model 16 that has been trained to receive the data as input and output information indicating changes in the state of the device 2, thereby outputting first estimated data (S3). In this estimation method, second data is input to the estimation model 16, thereby outputting second estimated data (S4). In this estimation method, the effect of the specific parameter on the state of the device 2 is estimated based on the relationship between the first estimated data and the second estimated data (S5).

[0059] This has the advantage that the first data and the second data, which differ only in the value of a specific parameter, are input into the same estimation model 16, making it easier to estimate the effect that a parameter (specific parameter) related to the state of the equipment 2 has on changes in the state of the equipment 2.

[0060] In addition, in the estimation method according to the second aspect, in the process of estimating the influence (S5) in the first aspect, the influence due to the difference between the value of a specific parameter in the first data and the value of a specific parameter in the second data is estimated based on the difference between the first estimated data and the second estimated data as the relationship between the first estimated data and the second estimated data (S51, S52).

[0061] This has the advantage that it becomes easier to accurately estimate how a difference in a specific parameter affects the state of the device 2.

[0062] In the estimation method according to the third aspect, in the first or second aspect, the specific parameter is a value related to an action of the user with respect to the device 2. The one or more parameters include at least one of a physical quantity representing a state of the device 2, a value related to control of the device 2, and a physical quantity representing a state around the device 2.

[0063] This has the advantage that it becomes easier to estimate the effect that a user's action on the device 2 has on the state of the device 2.

[0064] In addition, in the estimation method according to a fourth aspect, in any one of the first to third aspects, the device 2 is a refrigerator, and the change in the state of the device 2 is a change in the temperature inside the refrigerator.

[0065] This has the advantage that it becomes easier to estimate the effect of a specific parameter on changes in the temperature inside the refrigerator.

[0066] In addition, in the estimation method according to a fifth aspect, in the fourth aspect, the specific parameter is a feature based on at least one of the number of times the refrigerator door is opened and closed and the duration of the door opening and closing.

[0067] This has the advantage of making it easier to estimate the influence of at least one of the number of times the refrigerator door is opened and closed and the duration of the door opening and closing on the change in the temperature inside the refrigerator.

[0068] A program according to a sixth aspect causes one or more processors to execute the estimation method according to any one of the first to fifth aspects.

[0069] This has the advantage of making it possible to provide a program that makes it easy to estimate the influence that a parameter related to the state of the device 2 has on the state of the device 2.

[0070] Furthermore, an estimation system 1 according to a seventh aspect includes an acquisition unit 11, a generation unit 12, a first estimation unit 13, a second estimation unit 14, and a third estimation unit 15. The acquisition unit 11 acquires first data relating to the state of the device 2 and having a plurality of parameters including a specific parameter. The generation unit 12 generates second data by changing the value of the specific parameter in the first data acquired by the acquisition unit 11. The first estimation unit 13 generates first estimated data by inputting the first data acquired by the acquisition unit 11 into an estimation model 16 that is trained to receive the above data as input and output information indicating a change in the state of the device 2. The second estimation unit 14 generates second estimated data by inputting the second data generated by the generation unit 12 into the estimation model 16. The third estimation unit 15 estimates the effect of the specific parameter on the state of the device 2 based on the relationship between the first estimated data generated by the first estimation unit 13 and the second estimated data generated by the second estimation unit 14.

[0071] This has the advantage of providing an estimation system 1 that can easily estimate the influence of parameters related to the state of the device 2 on the device 2.

[0072] The present disclosure can be applied to a system for diagnosing the condition of a device such as a refrigerator.

[0073] REFERENCE SIGNS LIST 1 estimation system 11 acquisition unit 12 generation unit 13 first estimation unit 14 second estimation unit 15 third estimation unit 16 estimation model 2 device N1 network

Claims

1. An estimation method comprising: acquiring first data relating to a state of an equipment, the first data having a plurality of parameters including a specific parameter; generating second data by changing a value of the specific parameter in the acquired first data; outputting first estimated data by inputting the acquired first data into an estimation model trained to use the data as input and output information indicating a change in the state of the equipment; inputting the generated second data into the estimation model to output second estimated data; and estimating the effect of the specific parameter on the state of the equipment based on the relationship between the first estimated data and the second estimated data.

2. The estimation method according to claim 1, wherein in the process of estimating the influence, the influence due to the difference between the value of the specific parameter in the first data and the value of the specific parameter in the second data is estimated based on the difference between the first estimated data and the second estimated data as the relationship.

3. The estimation method according to claim 1 or 2, wherein the specific parameter is a value related to an action of a user with respect to the device, and the one or more parameters include at least one of a physical quantity representing a state of the device, a value related to control of the device, and a physical quantity representing a state around the device.

4. The estimation method according to claim 1 or 2, wherein the appliance is a refrigerator, and the change in the state of the appliance is a change in temperature inside the refrigerator.

5. The estimation method according to claim 4, wherein the specific parameter is a feature based on at least one of the number of times the refrigerator door is opened and closed and the duration of time the door is opened and closed.

6. A program causing one or more processors to execute the estimation method according to claim 1 or 2.

7. An estimation system comprising: an acquisition unit that acquires first data relating to a state of an equipment, the first data having a plurality of parameters including a specific parameter; a generation unit that generates second data by changing a value of the specific parameter in the first data acquired by the acquisition unit; a first estimation unit that outputs first estimated data by inputting the first data acquired by the acquisition unit into an estimation model that has been trained to use the data as input and output information indicating a change in the state of the equipment; a second estimation unit that outputs second estimated data by inputting the second data generated by the generation unit into the estimation model; and a third estimation unit that estimates an effect of the specific parameter on the state of the equipment based on a relationship between the first estimated data output by the first estimation unit and the second estimated data generated by the second estimation unit.

Images