Fault diagnosis method, model training method and device of air conditioner
By obtaining the operating data set of the air conditioner and determining the feature extraction method based on user operations and operation stages, the problems of air conditioner fault diagnosis accuracy and low efficiency are solved, and more efficient fault identification and equipment maintenance are achieved.
Patent Information
- Application Number
- CN202410958003.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-16
- Publication Date
- 2024-11-12
AI Technical Summary
The fault diagnosis of air conditioners is problematic of low accuracy and efficiency, and it is difficult to identify faults in a timely and accurate manner, affecting the normal operation of the equipment.
By acquiring the operating data set of the air conditioner, determining the target feature extraction method based on user operations and operation stages, feature extraction is performed on the data, and fault diagnosis is performed using the feature set, which improves the accuracy and efficiency of diagnosis.
It improves the accuracy and efficiency of air conditioner fault diagnosis, saves computing resources, can identify faults more accurately, and ensures the normal operation of the equipment.
Smart Images

Figure CN118940086A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the technical field of air conditioners, and in particular to a fault diagnosis method, a model training method, a device, an electronic device and a storage medium for an air conditioner. Background Art
[0002] At present, it is necessary to perform fault diagnosis on air conditioners, such as diagnosing faults such as dirtiness, blockage, and lack of fluorine, so as to timely understand whether the air conditioner has a fault, so as to perform maintenance on the air conditioner in time and ensure the normal operation of the air conditioner. However, in the related art, the fault diagnosis of air conditioners has the problems of low accuracy and efficiency. Summary of the invention
[0003] The present disclosure provides a fault diagnosis method, model training method, device, electronic device, computer-readable storage medium, and computer program product for an air conditioner, so as to at least solve the problem of low accuracy and efficiency in fault diagnosis of air conditioners in the related art. The technical solution of the present disclosure is as follows:
[0004] According to a first aspect of an embodiment of the present disclosure, there is provided a method for diagnosing faults of an air conditioner, comprising: obtaining an operating data set corresponding to the air conditioner, wherein the operating data set includes operating data of the air conditioner at multiple sampling moments; determining a target feature extraction method for the operating data set based on operations performed by a user on a target client corresponding to the air conditioner, and / or based on a target operating stage of the air conditioner corresponding to the operating data set; performing feature extraction on the operating data set according to the target feature extraction method to obtain an operating feature set corresponding to the air conditioner, wherein the operating feature set includes operating features of the air conditioner at multiple sampling moments; and performing fault diagnosis on the air conditioner based on the operating feature set.
[0005] In one embodiment of the present disclosure, the target feature extraction method of the operating data set is determined based on the user's operation on the target client corresponding to the air conditioner and / or based on the target operating stage of the air conditioner corresponding to the operating data set, including: controlling the display of multiple first candidate selection items on the interactive page of the target client, wherein the first candidate selection items are used to characterize the candidate feature extraction method; based on the user's selection operation on the interactive page for the first target selection item, using the candidate feature extraction method corresponding to the first target selection item as the target feature extraction method.
[0006] In one embodiment of the present disclosure, obtaining an operating data set corresponding to the air conditioner includes: controlling the display of multiple second candidate selection items and multiple third candidate selection items on the interactive page of the target client, wherein the second candidate selection items are used to represent candidate sampling time periods, and the third candidate selection items are used to represent candidate categories of the operating data of the air conditioner; based on the user's selection operation on the interactive page for the second target selection item, taking the candidate sampling time period corresponding to the second target selection item as the target sampling time period; based on the user's selection operation on the interactive page for the third target selection item, taking the candidate category corresponding to the third target selection item as the target category; and adding the operating data of the target category of the air conditioner under the target sampling time period to the operating data set.
[0007] In one embodiment of the present disclosure, before acquiring the operating data set corresponding to the air conditioner, the method further includes: detecting that the user has started a fault diagnosis function on the interactive page of the target client.
[0008] In one embodiment of the present disclosure, the target feature extraction method of the operating data set is determined based on the user's operation on the target client corresponding to the air conditioner, and / or based on the target operating stage of the air conditioner corresponding to the operating data set, including: determining the target feature category to be extracted from the operating data set based on the target operating stage as the target feature extraction method; performing feature extraction on the operating data set according to the target feature extraction method to obtain the operating feature set corresponding to the air conditioner, including: performing feature extraction on the operating data set to obtain the operating features of the target feature category; and adding the operating features of the target feature category to the operating feature set.
[0009] In one embodiment of the present disclosure, if there are multiple target operation stages, determining the target feature category to be extracted from the operation data set based on the target operation stage includes: splitting the operation data set to obtain multiple operation sub-data sets, wherein the operation sub-data sets correspond one-to-one to the target operation stage; determining the sub-feature category to be extracted from the operation sub-data set based on the target operation stage corresponding to the operation sub-data set; and performing feature extraction on the operation data set to obtain the operation feature of the target feature category includes: performing feature extraction on the operation sub-data set to obtain the operation feature of the sub-feature category.
[0010] In one embodiment of the present disclosure, determining the target feature category to be extracted from the operating data set based on the target operating stage includes: obtaining a mapping relationship between candidate operating stages of the air conditioner and candidate feature categories; obtaining a candidate feature category having a mapping relationship with the target operating stage as the target feature category.
[0011] In one embodiment of the present disclosure, the target feature category includes at least one of a dynamic feature category, a steady-state feature category, and a cumulative feature category.
[0012] In one embodiment of the present disclosure, determining the target feature category to be extracted from the operating data set based on the target operating stage includes: if the target operating stage includes a dynamic stage, determining that the target feature category includes the dynamic feature category and the cumulative feature category; if the target operating stage includes a steady-state stage, determining that the target feature category includes the steady-state feature category and the cumulative feature category.
[0013] In one embodiment of the present disclosure, the dynamic feature category includes a change speed and / or a change acceleration of the operating data of the air conditioner.
[0014] In one embodiment of the present disclosure, obtaining the target operating stage includes: if the operating data set includes operating data at a first moment, wherein multiple sampling moments include the first moment, and the cumulative operating time of the air conditioner at the first moment is less than or equal to the set time, determining that the target operating stage includes a dynamic stage; if the operating data set includes operating data at a second moment, wherein multiple sampling moments include the second moment, and the cumulative operating time of the air conditioner at the second moment is greater than the set time, determining that the target operating stage includes a steady-state stage.
[0015] In one embodiment of the present disclosure, obtaining the target operating stage includes: if the operating data set includes operating data at a third moment, wherein multiple sampling moments include the third moment, and the absolute value of the difference between the indoor ambient temperature and the set temperature at the third moment is greater than a set threshold, determining that the target operating stage includes a dynamic stage; if the operating data set includes operating data at a fourth moment, wherein multiple sampling moments include the fourth moment, and the absolute value of the difference at the fourth moment is less than or equal to the set threshold, determining that the target operating stage includes a steady-state stage.
[0016] In one embodiment of the present disclosure, the fault diagnosis of the air conditioner based on the operating feature set includes: inputting the operating feature set into a target fault diagnosis model, and the target fault diagnosis model outputs a fault diagnosis result of the air conditioner.
[0017] In one embodiment of the present disclosure, the method also includes: controlling the display of multiple fourth candidate selection items on the interactive page of the target client, wherein the fourth candidate selection items are used to characterize candidate fault diagnosis models; based on the user's selection operation on the fourth target selection item on the interactive page, the candidate fault diagnosis model corresponding to the fourth target selection item is used as the target fault diagnosis model.
[0018] In one embodiment of the present disclosure, the method further includes: controlling the display of fault diagnosis data of the air conditioner on the interactive page of the target client, wherein the fault diagnosis data includes the target feature extraction method and / or the fault diagnosis result of the air conditioner.
[0019] According to a second aspect of an embodiment of the present disclosure, a model training method is provided, comprising: obtaining a sample operation data set corresponding to a sample air conditioner, wherein the sample operation data set includes sample operation data of the sample air conditioner at multiple sampling moments; determining a sample feature extraction method of the sample operation data set based on operations of a sample user on a sample client corresponding to the sample air conditioner, and / or based on a sample operation stage of the sample air conditioner corresponding to the sample operation data set; performing feature extraction on the sample operation data set according to the sample feature extraction method to obtain a sample operation feature set corresponding to the sample air conditioner, wherein the sample operation feature set includes sample operation features of the sample air conditioner at multiple sampling moments; and training a fault diagnosis model based on the sample operation feature set.
[0020] In one embodiment of the present disclosure, the sample feature extraction method of the sample operation data set is determined based on the operation of the sample user on the sample client corresponding to the sample air conditioner, and / or based on the sample operation stage of the sample air conditioner corresponding to the sample operation data set, including: controlling the display of multiple first candidate selection items on the interactive page of the sample client, wherein the first candidate selection item is used to characterize the candidate feature extraction method; based on the selection operation of the sample user on the interactive page for the first target selection item, using the candidate feature extraction method corresponding to the first target selection item as the sample feature extraction method.
[0021] In one embodiment of the present disclosure, the method of extracting sample features of the sample operation data set based on the operation of the sample user on the sample client corresponding to the sample air conditioner and / or based on the sample operation stage of the sample air conditioner corresponding to the sample operation data set includes: determining, based on the sample operation stage, a sample feature category to be extracted from the sample operation data set as the sample feature extraction method;
[0022] The method of performing feature extraction on the sample operation data set according to the sample feature extraction method to obtain a sample operation feature set corresponding to the sample air conditioner includes: performing feature extraction on the sample operation data set to obtain sample operation features of the sample feature category; and adding the sample operation features of the sample feature category to the sample operation feature set.
[0023] In one embodiment of the present disclosure, the sample operation data set also includes sample fault diagnosis results of the sample air conditioner, and the training of the fault diagnosis model based on the sample operation feature set includes: inputting the sample operation feature set into the fault diagnosis model, and the fault diagnosis model outputs the predicted fault diagnosis results of the sample air conditioner; and training the fault diagnosis model based on the predicted fault diagnosis results of the sample air conditioner and the sample fault diagnosis results of the sample air conditioner.
[0024] According to a third aspect of an embodiment of the present disclosure, there is provided a fault diagnosis device for an air conditioner, comprising: an acquisition module, configured to execute acquisition of an operation data set corresponding to the air conditioner, wherein the operation data set includes operation data of the air conditioner at multiple sampling moments; a determination module, configured to execute a target feature extraction method for the operation data set based on operations of a user on a target client corresponding to the air conditioner, and / or based on a target operation stage of the air conditioner corresponding to the operation data set; a feature extraction module, configured to execute feature extraction on the operation data set according to the target feature extraction method to obtain an operation feature set corresponding to the air conditioner, wherein the operation feature set includes operation features of the air conditioner at multiple sampling moments; and a fault diagnosis module, configured to perform fault diagnosis on the air conditioner based on the operation feature set.
[0025] In one embodiment of the present disclosure, the determination module is further configured to execute: controlling the display of multiple first candidate selection items on the interactive page of the target client, wherein the first candidate selection items are used to represent candidate feature extraction methods; based on the user's selection operation on the interactive page for the first target selection item, using the candidate feature extraction method corresponding to the first target selection item as the target feature extraction method.
[0026] In one embodiment of the present disclosure, the acquisition module is further configured to execute: controlling the display of multiple second candidate selection items and multiple third candidate selection items on the interactive page of the target client, wherein the second candidate selection items are used to represent candidate sampling time periods, and the third candidate selection items are used to represent candidate categories of the operating data of the air conditioner; based on the user's selection operation on the second target selection item on the interactive page, taking the candidate sampling time period corresponding to the second target selection item as the target sampling time period; based on the user's selection operation on the third target selection item on the interactive page, taking the candidate category corresponding to the third target selection item as the target category; and adding the operating data of the target category of the air conditioner under the target sampling time period to the operating data set.
[0027] In one embodiment of the present disclosure, before acquiring the operating data set corresponding to the air conditioner, the acquisition module is further configured to execute: detecting the user's activation operation for the fault diagnosis function on the interactive page of the target client.
[0028] In one embodiment of the present disclosure, the determination module is further configured to execute: based on the target operation stage, determining the target feature category to be extracted from the operation data set as the target feature extraction method;
[0029] The feature extraction module is further configured to execute: extracting features from the operation data set to obtain operation features of the target feature category; and adding the operation features of the target feature category to the operation feature set.
[0030] In one embodiment of the present disclosure, the determining module is further configured to execute: if there are multiple target operation stages, split the operation data set to obtain multiple operation sub-data sets, wherein the operation sub-data sets correspond one-to-one to the target operation stages;
[0031] The feature extraction module is further configured to execute: obtaining a target feature category determined by a target operation stage corresponding to the operation sub-dataset as a sub-feature category to be extracted from the operation sub-dataset; and performing feature extraction on the operation sub-dataset to obtain operation features of the sub-feature category.
[0032] In one embodiment of the present disclosure, the determination module is further configured to execute: obtaining a mapping relationship between candidate operating stages of the air conditioner and candidate feature categories; obtaining a candidate feature category having a mapping relationship with the target operating stage as the target feature category.
[0033] In one embodiment of the present disclosure, the target feature category includes at least one of a dynamic feature category, a steady-state feature category, and a cumulative feature category.
[0034] In one embodiment of the present disclosure, the determination module is further configured to execute: if the target operation stage includes a dynamic stage, determining that the target feature category includes the dynamic feature category and the cumulative feature category; if the target operation stage includes a steady-state stage, determining that the target feature category includes the steady-state feature category and the cumulative feature category.
[0035] In one embodiment of the present disclosure, the dynamic feature category includes a change speed and / or a change acceleration of the operating data of the air conditioner.
[0036] In one embodiment of the present disclosure, the determination module is further configured to execute: if the operating data set includes operating data at a first moment, wherein multiple sampling moments include the first moment, and the cumulative operating time of the air conditioner at the first moment is less than or equal to the set time, it is determined that the target operating stage includes a dynamic stage; if the operating data set includes operating data at a second moment, wherein multiple sampling moments include the second moment, and the cumulative operating time of the air conditioner at the second moment is greater than the set time, it is determined that the target operating stage includes a steady-state stage.
[0037] In one embodiment of the present disclosure, the determination module is further configured to execute: if the operating data set includes operating data at a third moment, wherein multiple sampling moments include the third moment, and the absolute value of the difference between the indoor ambient temperature and the set temperature at the third moment is greater than the set threshold, it is determined that the target operating stage includes a dynamic stage; if the operating data set includes operating data at a fourth moment, wherein multiple sampling moments include the fourth moment, and the absolute value of the difference at the fourth moment is less than or equal to the set threshold, it is determined that the target operating stage includes a steady-state stage.
[0038] In one embodiment of the present disclosure, the fault diagnosis module is further configured to execute: inputting the operating feature set into a target fault diagnosis model, and having the target fault diagnosis model output a fault diagnosis result of the air conditioner.
[0039] In one embodiment of the present disclosure, the fault diagnosis module is further configured to execute: controlling the display of multiple fourth candidate selection items on the interactive page of the target client, wherein the fourth candidate selection items are used to characterize candidate fault diagnosis models; based on the user's selection operation on the fourth target selection item on the interactive page, using the candidate fault diagnosis model corresponding to the fourth target selection item as the target fault diagnosis model.
[0040] In one embodiment of the present disclosure, the fault diagnosis module is further configured to execute: controlling the display of fault diagnosis data of the air conditioner on the interactive page of the target client, wherein the fault diagnosis data includes the target feature extraction method and / or the fault diagnosis result of the air conditioner.
[0041] According to a fourth aspect of an embodiment of the present disclosure, a model training device is provided, comprising: an acquisition module, configured to execute acquisition of a sample operation data set corresponding to a sample air conditioner, wherein the sample operation data set includes sample operation data of the sample air conditioner at multiple sampling moments; a determination module, configured to execute a sample feature extraction method of the sample operation data set based on operations of a sample user on a sample client corresponding to the sample air conditioner, and / or based on a sample operation stage of the sample air conditioner corresponding to the sample operation data set; a feature extraction module, configured to execute feature extraction on the sample operation data set according to the sample feature extraction method to obtain a sample operation feature set corresponding to the sample air conditioner, wherein the sample operation feature set includes sample operation features of the sample air conditioner at multiple sampling moments; and a training module, configured to execute training of a fault diagnosis model based on the sample operation feature set.
[0042] In one embodiment of the present disclosure, the determination module is further configured to execute: controlling the display of multiple first candidate selection items on the interactive page of the sample client, wherein the first candidate selection items are used to characterize candidate feature extraction methods; based on the selection operation of the sample user on the interactive page for the first target selection item, using the candidate feature extraction method corresponding to the first target selection item as the sample feature extraction method.
[0043] In one embodiment of the present disclosure, the determination module is further configured to execute: based on the sample operation stage, determining the sample feature category to be extracted from the sample operation data set as the sample feature extraction method;
[0044] The feature extraction module is further configured to execute: extracting features from the sample operation data set to obtain sample operation features of the sample feature category; and adding the sample operation features of the sample feature category to the sample operation feature set.
[0045] In one embodiment of the present disclosure, the sample operation data set also includes sample fault diagnosis results of the sample air conditioner, and the training module is further configured to execute: inputting the sample operation feature set into the fault diagnosis model, and the fault diagnosis model outputs the predicted fault diagnosis results of the sample air conditioner; and training the fault diagnosis model based on the predicted fault diagnosis results of the sample air conditioner and the sample fault diagnosis results of the sample air conditioner.
[0046] According to a fifth aspect of an embodiment of the present disclosure, an electronic device is provided, comprising a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to implement the steps of the method described in the first aspect and the second aspect of the embodiment of the present disclosure.
[0047] According to a sixth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, on which computer program instructions are stored. When the program instructions are executed by a processor, the steps of the method described in the first aspect and the second aspect of the embodiment of the present disclosure are implemented.
[0048] According to the seventh aspect of the embodiments of the present disclosure, a computer program product is provided, including a computer program, characterized in that when the computer program is executed by a processor of an electronic device, the steps of the method described in the first aspect and the second aspect of the embodiments of the present disclosure are implemented.
[0049] The technical solution provided by the embodiment of the present disclosure brings at least the following beneficial effects: obtaining an operation data set corresponding to the air conditioner, wherein the operation data set includes the operation data of the air conditioner at multiple sampling moments, determining a target feature extraction method for the operation data set based on the user's operation on the target client corresponding to the air conditioner, and / or based on the target operation stage of the air conditioner corresponding to the operation data set, performing feature extraction on the operation data set according to the target feature extraction method, and obtaining an operation feature set corresponding to the air conditioner, wherein the operation feature set includes the operation features of the air conditioner at multiple sampling moments, and performing fault diagnosis on the air conditioner based on the operation feature set. Thus, the target feature extraction method can be determined to extract features from the operation data set in consideration of the user's operation on the interactive page and / or the target operation stage of the air conditioner corresponding to the operation data set, thereby improving the accuracy of the target feature extraction method, facilitating the collection of features that are relatively small in number and strongly related to the fault diagnosis of the air conditioner, and automatically performing fault diagnosis on the air conditioner using the operation feature set, thereby improving the accuracy and efficiency of the fault diagnosis of the air conditioner and saving computing resources.
[0050] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute improper limitations on the present disclosure.
[0052] Figure 1 The figure is a flow chart showing a method for diagnosing a fault of an air conditioner according to an exemplary embodiment.
[0053] Figure 2 is a flow chart showing a fault diagnosis method for an air conditioner according to another exemplary embodiment.
[0054] Figure 3 is a flow chart showing a fault diagnosis method for an air conditioner according to another exemplary embodiment.
[0055] Figure 4 The figure is a flowchart of a model training method according to an exemplary embodiment.
[0056] Figure 5 The invention is a block diagram of a fault diagnosis device for an air conditioner according to an exemplary embodiment.
[0057] Figure 6 It is a block diagram of a model training device according to an exemplary embodiment.
[0058] Figure 7 It is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION
[0059] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings.
[0060] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.
[0061] The acquisition, storage, use, and processing of data in the technical solution disclosed in this disclosure are in compliance with the relevant laws and regulations.
[0062] This disclosure is based on the analysis of the heat conduction process principle of the air conditioner and the understanding of the operation process of the air conditioner. The energy generated by the air conditioner compressor is transferred to the interior of the room through the heat conduction process of the refrigerant, evaporator, air and other media. In this process, the compressor, as the "heart" of the air conditioner, plays a core role. fan , Outdoor ambient temperature T o (t), indoor ambient temperature T i (t), set temperature, exhaust temperature T e (t), inner tube temperature T g (t), external machine energy consumption power P, external machine cooling power Q c There is a dependency relationship between the variables. In some scenarios, the indoor fan speed v fan Remain unchanged. When the air conditioner is in normal or abnormal state, the relationship between the state of each component of the air conditioner and the compressor frequency freq is as follows:
[0063] P = f1(freq)
[0064] Q c =f2(freq)
[0065] T e (t) = f3(freq)
[0066]
[0067] It should be noted that the first function f1 is a function with the compressor frequency freq as the independent variable and the outdoor machine energy consumption power P as the dependent variable, and the second function f2 is a function with the compressor frequency freq as the independent variable and the outdoor machine refrigeration power Q c The third function f3 is a function of the dependent variable, with the compressor frequency freq as the independent variable and the exhaust temperature T e (t) is the function of the dependent variable, C g is the equivalent specific heat capacity of the indoor tube, R g is the equivalent thermal resistance of the indoor pipe, Q c is negative, the air conditioner is in the heating stage and the lower Q c is positive, C i is the equivalent specific heat capacity of indoor air, R i is the equivalent thermal resistance of indoor air, R w is the equivalent thermal resistance of the room wall.
[0068] For example, the second function f2 and the third function f3 are both piecewise linear functions.
[0069] Figure 1 is a flow chart showing a method for diagnosing a fault of an air conditioner according to an exemplary embodiment. Figure 1As shown, the fault diagnosis method of the air conditioner according to the embodiment of the present disclosure includes the following steps.
[0070] S101, obtaining an operation data set corresponding to an air conditioner, wherein the operation data set includes operation data of the air conditioner at multiple sampling moments.
[0071] It should be noted that the executor of the fault diagnosis method for the air conditioner of the embodiment of the present disclosure is an electronic device, such as a mobile phone, a notebook, a desktop computer, a vehicle-mounted terminal, a smart home appliance, a wearable device, etc. Among them, the wearable device may include a wrist-worn device (such as a smart watch, a smart bracelet), a head-worn device, a foot-worn device, etc. The fault diagnosis method for the air conditioner of the embodiment of the present disclosure can be executed by the fault diagnosis device for the air conditioner of the embodiment of the present disclosure, and the fault diagnosis device for the air conditioner of the embodiment of the present disclosure can be configured in any electronic device to execute the fault diagnosis method for the air conditioner of the embodiment of the present disclosure.
[0072] It should be noted that there are no excessive restrictions on air conditioners, for example, they may include refrigeration air conditioners, heating air conditioners, etc. There are no excessive restrictions on operating data, for example, they may include indoor ambient temperature, outdoor ambient temperature, compressor frequency, exhaust temperature, inner pipe temperature, outdoor unit power consumption, indoor fan speed, etc. Among them, the inner pipe temperature refers to the pipe wall temperature of the indoor heat exchanger.
[0073] It should be noted that the time interval between two adjacent sampling moments is the sampling period, and there is no excessive limitation on the sampling period, for example, it may include 5 minutes, 10 minutes, etc.
[0074] For example, the operation data set A includes the operation data of the air conditioner at sampling times 1 to 7, where sampling time 1 is the start-up time of the air conditioner, sampling time 2 is the time when the cumulative operation time of the air conditioner is 10 minutes, sampling time 3 is the time when the cumulative operation time of the air conditioner is 20 minutes, sampling time 4 is the time when the cumulative operation time of the air conditioner is 30 minutes, sampling time 5 is the time when the cumulative operation time of the air conditioner is 40 minutes, sampling time 6 is the time when the cumulative operation time of the air conditioner is 50 minutes, and sampling time 7 is the time when the cumulative operation time of the air conditioner is 1 hour. Among them, the cumulative operation time of the air conditioner is the cumulative time counted from the start-up time of the air conditioner. It can be understood that the sampling time period is the time period between sampling times 1 to 7, that is, the time period when the air conditioner runs continuously for 1 hour from the start-up time, the sampling time is 1 hour, and the sampling period is 10 minutes.
[0075] For example, the operation data set B includes the operation data of the air conditioner at sampling times 1 to 4. The relevant contents of sampling times 1 to 4 can be found in the above embodiment, and will not be described again here.
[0076] For example, the operation data set C includes the operation data of the air conditioner at sampling times 5 to 7. The relevant contents of sampling times 5 to 7 can be found in the above embodiment, and will not be described again here.
[0077] In one embodiment, obtaining an operation data set corresponding to an air conditioner includes controlling to display a plurality of second candidate selection items and a plurality of third candidate selection items on an interactive page of a target client, wherein the second candidate selection item is used to characterize a candidate sampling time period, and the third candidate selection item is used to characterize a candidate category of the operation data of the air conditioner, based on a selection operation of a user on the interactive page for the second target selection item, taking the candidate sampling time period corresponding to the second target selection item as a target sampling time period, and taking the candidate category corresponding to the third target selection item as a target category based on a selection operation of a user on the interactive page for the third target selection item, and adding the operation data of the target category of the air conditioner in the target sampling time period to the operation data set. Thus, the candidate sampling time period selected by the user can be taken as the target sampling time period, and the candidate category of the operation data selected by the user can be taken as the target category to generate the operation data set, and personalized generation of the operation data set can be achieved.
[0078] It should be noted that the target client is the client corresponding to the air conditioner, and there are no excessive restrictions on users and target clients. For example, users may include maintenance personnel and users of the air conditioner, and target clients may include clients of maintenance personnel and users of the air conditioner. Multiple second candidate options include a second target option, and multiple third candidate options include a third target option. The number of second target options, third target options, target sampling time periods, and target categories is at least one. There are no excessive restrictions on selection operations, for example, they may include click operations, drag operations, etc. There are no excessive restrictions on the display methods of candidate options, for example, they may include pop-up window display, notification bar display, floating window display, etc.
[0079] For example, the second candidate selection items 1 to 3 may be displayed on the interactive page of the target client, wherein the second candidate selection item 1 is used to represent June 1 to June 30, the second candidate selection item 2 is used to represent July 1 to July 31, and the second candidate selection item 3 is used to represent August 1 to August 31. Among them, June 1 to June 30, July 1 to July 31, and August 1 to August 31 are all candidate sampling time periods.
[0080] For example, based on the user's selection operation on the interactive page for the second candidate option 1, June 1 to June 30 may be used as the target sampling period. It should be noted that in this embodiment, the second candidate option 1 is the second target option.
[0081] For example, third candidate selection items 1 to 7 may be displayed on the interactive page of the target client, wherein third candidate selection item 1 is used to characterize the indoor ambient temperature, third candidate selection item 2 is used to characterize the outdoor ambient temperature, third candidate selection item 3 is used to characterize the compressor frequency, third candidate selection item 4 is used to characterize the exhaust temperature, third candidate selection item 5 is used to characterize the inner pipe temperature, third candidate selection item 6 is used to characterize the energy consumption power of the outdoor unit, and third candidate selection item 7 is used to characterize the indoor fan speed. Among them, the indoor ambient temperature, outdoor ambient temperature, compressor frequency, exhaust temperature, inner pipe temperature, energy consumption power of the outdoor unit, and indoor fan speed are all candidate categories of operation data.
[0082] For example, based on the user's selection operation on the interactive page for the third candidate selection items 1 to 7, the indoor ambient temperature, outdoor ambient temperature, compressor frequency, exhaust temperature, internal pipe temperature, outdoor unit energy consumption power, and indoor fan speed can all be used as target categories. It should be noted that in this embodiment, the third candidate selection items 1 to 7 are all third target selection items.
[0083] In one embodiment, before obtaining the operating data set corresponding to the air conditioner, it also includes detecting the user's activation operation for the fault diagnosis function on the interactive page of the target client. Therefore, when the user activates the fault diagnosis function, step S101 and subsequent steps are continued to perform fault diagnosis on the air conditioner.
[0084] It should be noted that there are no excessive restrictions on the activation operation, for example, it may include touch operation, gesture operation, posture operation, etc. For example, a button for activating the fault diagnosis function may be displayed on the interactive page of the target client, and a user's click operation on the button for activating the fault diagnosis function on the interactive page may be detected as an activation operation for the fault diagnosis function. For example, a user's setting gesture operation and / or setting posture operation on the interactive page may be detected as an activation operation for the fault diagnosis function.
[0085] S102, determining a target feature extraction method for the operation data set based on the user's operation on the target client corresponding to the air conditioner and / or based on the target operation stage of the air conditioner corresponding to the operation data set.
[0086] It should be noted that the target operation stage refers to the operation stage of the air conditioner at the sampling time corresponding to the operation data in the operation data set. The number of target operation stages is at least one. If the target operation stage is one, it indicates that at the sampling time corresponding to each operation data in the operation data set, the air conditioner is in the same operation stage. Conversely, if the target operation stage is multiple, it indicates that at the sampling time corresponding to at least two operation data in the operation data set, the air conditioner is in different operation stages. Different operation data sets may correspond to different target operation stages. For the relevant content of obtaining the target operation stage, please refer to the following embodiments, which will not be repeated here.
[0087] It should be noted that there are no excessive restrictions on the target operation stage, for example, it may include a dynamic stage, a steady-state stage, a cooling stage, a heating stage, a defrosting stage, a cleaning stage, a dehumidification stage, a ventilation stage, etc. Among them, some operating parameters of the air conditioner in the dynamic stage (such as indoor ambient temperature, compressor frequency, exhaust temperature, internal pipe temperature, and outdoor unit power consumption) vary greatly, while some operating parameters of the air conditioner in the steady-state stage vary less, that is, they remain at the set value.
[0088] In one embodiment, the dynamic stage is a stage in which the indoor ambient temperature changes from the initial temperature to the set temperature, and the steady-state stage is a stage in which the indoor ambient temperature is maintained at the set temperature. It is understood that the absolute value of the difference between the indoor ambient temperature and the set temperature in the dynamic stage is greater than the set threshold, and the absolute value of the difference between the indoor ambient temperature and the set temperature in the steady-state stage is less than or equal to the set threshold. The initial temperature is the initial value of the indoor ambient temperature, and the set temperature is the target value of the indoor ambient temperature. The set temperature can be set manually or by an algorithm, and there are no excessive restrictions on the set temperature and the set threshold. For example, the set temperature can be 22°C, and the set threshold can be 0.5°C.
[0089] In one embodiment, the dynamic stage is a stage in which the accumulated operation time of the air conditioner is less than or equal to the set time, and the steady stage is a stage in which the accumulated operation time of the air conditioner is greater than the set time. There is no excessive limitation on the set time, for example, it can include 30 minutes.
[0090] It should be noted that there are not too many restrictions on the target feature extraction method. For example, it may include the target feature category to be extracted from the running data set, the number of features to be extracted, etc.
[0091] In the embodiments of the present disclosure, based on the user's operation on the target client corresponding to the air conditioner and / or based on the target operation stage of the air conditioner corresponding to the operation data set, the target feature extraction method of the operation data set is determined, including the following possible implementation methods:
[0092] Method 1: Control the display of multiple first candidate selection items on the interactive page of the target client, wherein the first candidate selection item is used to represent the candidate feature extraction method, and based on the user's selection operation on the interactive page for the first target selection item, the candidate feature extraction method corresponding to the first target selection item is used as the target feature extraction method.
[0093] Therefore, the candidate feature extraction method selected by the user can be used as the target feature extraction method, so that personalized acquisition of the target feature extraction method can be achieved.
[0094] It should be noted that the multiple first candidate options include a first target option, and the number of the first target option and the number of the target feature extraction methods are both at least one.
[0095] For example, the first candidate selection items 1 to 3 may be displayed on the interactive page of the target client, wherein the first candidate selection item 1 is used to represent the dynamic feature category, the first candidate selection item 2 is used to represent the steady-state feature category, and the first candidate selection item 3 is used to represent the cumulative feature category. The dynamic feature category, the steady-state feature category, and the cumulative feature category are all candidate feature categories for the running feature. It should be noted that the relevant contents of the dynamic feature category, the steady-state feature category, and the cumulative feature category can be found in the following embodiments, which will not be repeated here.
[0096] For example, both the dynamic feature category and the steady-state feature category can be used as the target feature category based on the user's selection operation on the first candidate selection items 1 and 2 on the interactive page. It should be noted that in this embodiment, the first candidate selection items 1 and 2 are both first target selection items.
[0097] Method 2: Based on the target operation stage, determine the target screening method of the operation data set, determine the target feature extraction method, including screening the target data set from the operation data set according to the target screening method, and extract features from the target data set to obtain an operation feature set.
[0098] Therefore, the target operation stage can be taken into consideration to determine the target screening method of the operation data set, so as to screen out the target data set from the operation data set, and perform feature extraction on the target data set to obtain an operation feature set.
[0099] It should be noted that the target dataset is a subset of the running dataset.
[0100] In one embodiment, based on the target operation stage, determining a target screening mode of the operation data set includes determining, based on the target operation stage, a target data category to be screened for the operation data set as a target screening mode. According to the target screening mode, screening out the target data set from the operation data set includes screening out operation data of the target data category from the operation data set, and adding the operation data of the target data category to the target data set.
[0101] In some examples, based on the target operation stage, determining the target data category to be screened for the operation data set includes obtaining a mapping relationship between the candidate operation stages of the air conditioner and the candidate data categories, and obtaining the candidate data category having a mapping relationship with the target operation stage as the target data category. It is understandable that different candidate operation stages may be mapped to different candidate data categories.
[0102] Method 3: If there are multiple target operation stages, determining the target feature extraction method includes splitting the operation data set to obtain multiple operation sub-data sets, wherein the operation sub-data sets correspond one-to-one to the target operation stages, performing feature extraction on the operation sub-data sets to obtain an operation sub-feature set, and obtaining a collection of the multiple operation sub-feature sets as the operation feature set.
[0103] Therefore, when there are multiple target operation stages, the operation data set can be split according to the target operation stage to obtain multiple operation sub-data sets, and feature extraction is performed on each operation sub-data set to obtain multiple operation sub-feature sets to obtain an operation feature set.
[0104] It should be noted that at the sampling time corresponding to each operation data in the operation sub-data set, the air conditioner is in the same operation stage, which is the target operation stage corresponding to the operation sub-data set.
[0105] For example, continuing to take the operation data set A as an example, the target operation stage of the air conditioner corresponding to the operation data set A includes the dynamic stage and the steady-state stage. The operation data set A can be split to obtain operation sub-datasets A1 and A2. The operation sub-dataset A1 includes the operation data of the air conditioner at sampling times 1 to 4, and the operation sub-dataset A2 includes the operation data of the air conditioner at sampling times 5 to 7. The target operation stage corresponding to the operation sub-dataset A1 is the dynamic stage, and the target operation stage corresponding to the operation sub-dataset A2 is the steady-state stage.
[0106] Feature extraction can be performed on the operation sub-dataset A1 to obtain the operation sub-feature set D1, and feature extraction can be performed on the operation sub-dataset A2 to obtain the operation sub-feature set D2. The collection of the operation sub-feature sets D1 and D2 is obtained as the operation feature set D.
[0107] In one embodiment, the operation data set is split to obtain multiple operation sub-data sets, including if the target operation stages corresponding to the multiple operation data in the operation data set are the same, the multiple operation data are divided into the same operation sub-data set. It should be noted that the target operation stage corresponding to the operation data refers to the operation stage of the air conditioner at the sampling time corresponding to the operation data.
[0108] S103, performing feature extraction on the operation data set according to the target feature extraction method to obtain an operation feature set corresponding to the air conditioner, wherein the operation feature set includes operation features of the air conditioner at multiple sampling moments.
[0109] It should be noted that there are no excessive restrictions on the categories of operating characteristics. For example, they may include at least one of dynamic characteristic categories, steady-state characteristic categories, and cumulative characteristic categories. Among them, the dynamic characteristic category may include the statistical value of the operating data (such as the value, average value, variance, mode, maximum value, minimum value, etc. of the operating data), the change speed and / or change acceleration of the operating data, the steady-state characteristic category may include the statistical value of the operating data, and the cumulative characteristic category may include the cumulative change amount of the operating data. The categories of operating characteristics of different operating data may be different.
[0110] In one embodiment, the dynamic feature category at the sampling moment may include the value of the operating data at the sampling moment, and the average value, variance, mode, maximum value, minimum value, etc. of multiple operating data in the time period where the sampling moment is located, and the change speed and / or change acceleration of the operating data at the sampling moment. For example, the time period where the sampling moment is located may include the time period between 5 seconds before the sampling moment and 5 seconds after the sampling moment.
[0111] In one embodiment, the steady-state feature category at the sampling moment may include the value of the operating data at the sampling moment, and the average value, variance, mode, maximum value, minimum value, etc. of multiple operating data within the time period of the sampling moment.
[0112] In one implementation, the cumulative feature category at the sampling time may include the cumulative change in the operating data from the start-up time of the air conditioner to the sampling time.
[0113] S104, performing fault diagnosis on the air conditioner based on the operation feature set.
[0114] It should be noted that the fault diagnosis of the air conditioner based on the operating feature set can be implemented by using any air conditioner fault diagnosis method in the relevant technology, and no excessive restrictions are made here.
[0115] In one embodiment, fault diagnosis is performed on the air conditioner based on an operating feature set, including identifying whether set conditions of a target fault diagnosis result are met based on the operating feature set; if the set conditions of the target fault diagnosis result are met, using the target fault diagnosis result as the fault diagnosis result of the air conditioner.
[0116] It should be noted that there are no excessive restrictions on the fault diagnosis results. For example, it may include whether there is a fault (such as normal, dirty, blocked, or lack of fluorine), the fault category (such as dirty, blocked, or lack of fluorine), the fault probability (such as the fault probability under the fault category), the fault level, the risk score, etc. The target fault diagnosis result is a partial candidate fault diagnosis result. There are no excessive restrictions on the setting conditions of the candidate fault diagnosis results. For example, it may include that the statistical value of the operating feature is greater than the set value, the statistical value of the operating feature is less than or equal to the set value, etc. The setting conditions of different candidate fault diagnosis results may be different.
[0117] In some examples, obtaining the setting conditions of the target fault diagnosis result includes obtaining a mapping relationship between the candidate fault diagnosis results and the setting conditions, and obtaining the setting conditions having a mapping relationship with the target fault diagnosis result as the setting conditions of the target fault diagnosis result. It is understandable that the setting conditions mapped to different candidate fault diagnosis results may be different.
[0118] In one embodiment, based on the operation feature set, the fault diagnosis of the air conditioner is performed, including inputting the operation feature set into a target fault diagnosis model, and the target fault diagnosis model outputs the fault diagnosis result of the air conditioner. Thus, the operation feature set can be processed by the target fault diagnosis model to perform fault diagnosis on the air conditioner.
[0119] It should be noted that there are not too many restrictions on the target fault diagnosis model. For example, it can include logistic regression model, decision tree model, DNN (Deep Neural Network) model, LGBM (Light Gradient Boosting Machine), Xgboost (EXtreme Gradient Boosting) model, etc.
[0120] For example, the operating feature set of air conditioner A can be input into the target fault diagnosis model, and the target fault diagnosis model outputs the probability P(y=0)=0.85 that air conditioner A is operating normally, the probability P(y=1)=0.12 that air conditioner A is dirty and blocked, and the probability P(y=2)=0.03 that air conditioner A is lacking fluorine, and outputs the fault diagnosis result that air conditioner A is operating normally.
[0121] For example, the operating feature set of air conditioner B can be input into the target fault diagnosis model, and the target fault diagnosis model outputs the probability P(y=0)=0.21 that air conditioner B is operating normally, the probability P(y=1)=0.69 that air conditioner B is dirty or clogged, and the probability P(y=2)=0.10 that air conditioner B is short of fluorine, and outputs the fault diagnosis result that air conditioner B is dirty or clogged.
[0122] For example, the operating feature set of air conditioner C can be input into the target fault diagnosis model, and the target fault diagnosis model outputs the probability P(y=0)=0.15 that air conditioner C is operating normally, the probability P(y=1)=0.32 that air conditioner C is dirty or clogged, and the probability P(y=2)=0.53 that air conditioner C is short of fluorine, and outputs the fault diagnosis result of air conditioner C being dirty or clogged.
[0123] In some examples, the method further includes controlling to display a plurality of fourth candidate selection items on an interactive page of the target client, wherein the fourth candidate selection item is used to characterize the candidate fault diagnosis model, and based on the user's selection operation on the interactive page for the fourth target selection item, the candidate fault diagnosis model corresponding to the fourth target selection item is used as the target fault diagnosis model. Thus, the candidate fault diagnosis model selected by the user can be used as the target fault diagnosis model, and personalized acquisition of the target fault diagnosis model can be achieved.
[0124] In one embodiment, after the air conditioner is fault-diagnosed based on the operating feature set, if the fault diagnosis result of the air conditioner indicates that the air conditioner is faulty, a maintenance work order is generated for manual review of the air conditioner, and the maintenance work order is sent to the maintenance personnel's client so that the maintenance personnel can manually review the fault diagnosis result of the air conditioner and correct any errors in the online fault diagnosis result in a timely manner.
[0125] In one embodiment, the method also includes controlling the display of fault diagnosis data of the air conditioner on an interactive page of the target client, wherein the fault diagnosis data includes a target feature extraction method and / or a fault diagnosis result of the air conditioner, and promptly informing a user who controls the target client of the fault diagnosis data of the air conditioner (such as a target feature extraction method and / or a fault diagnosis result of the air conditioner), that is, explicitly displaying the fault diagnosis data of the air conditioner to the user. In addition, when the fault diagnosis result of the air conditioner indicates that the air conditioner has a fault, the user can use the fault diagnosis data of the air conditioner as a repair certificate to handle fault repair and other services, thereby reducing the difficulty of obtaining evidence for users to report air conditioner faults and improving user experience.
[0126] It should be noted that fault diagnosis data refers to data generated during the air conditioner fault diagnosis process, and there are no excessive restrictions on fault diagnosis data. For example, it may include the description text of each step in the air conditioner fault diagnosis process, the intermediate results generated during the fault diagnosis process (such as operating data sets, operating feature sets), feature categories for feature extraction of operating data sets, target feature extraction methods, air conditioner fault diagnosis results, etc. There are no excessive restrictions on the display method of fault diagnosis data, such as pop-up window display, notification bar display, floating window display, etc.
[0127] The fault diagnosis method of the air conditioner provided by the embodiment of the present disclosure obtains the operation data set corresponding to the air conditioner, wherein the operation data set includes the operation data of the air conditioner at multiple sampling moments, determines the target feature extraction method of the operation data set based on the user's operation on the target client corresponding to the air conditioner, and / or based on the target operation stage of the air conditioner corresponding to the operation data set, performs feature extraction on the operation data set according to the target feature extraction method, obtains the operation feature set corresponding to the air conditioner, wherein the operation feature set includes the operation features of the air conditioner at multiple sampling moments, and performs fault diagnosis on the air conditioner based on the operation feature set. Thus, the target feature extraction method can be determined to extract features from the operation data set in consideration of the user's operation on the interactive page and / or the target operation stage of the air conditioner corresponding to the operation data set, thereby improving the accuracy of the target feature extraction method, facilitating the collection of features that are relatively small in number and strongly related to the fault diagnosis of the air conditioner, and automatically performing fault diagnosis on the air conditioner using the operation feature set, thereby improving the accuracy and efficiency of the fault diagnosis of the air conditioner and saving computing resources.
[0128] On the basis of any of the above embodiments, the target operating stage can be obtained by using any method for obtaining the operating stage of an air conditioner in the related art, which is not limited here.
[0129] In the embodiments of the present disclosure, obtaining the target operation stage may include the following possible implementation methods:
[0130] Method 1: If the operating data set includes operating data at a first moment, wherein multiple sampling moments include the first moment, and the cumulative operating time of the air conditioner at the first moment is less than or equal to the set time, it is determined that the target operating stage includes a dynamic stage; if the operating data set includes operating data at a second moment, wherein multiple sampling moments include the second moment, and the cumulative operating time of the air conditioner at the second moment is greater than the set time, it is determined that the target operating stage includes a steady-state stage.
[0131] It can be understood that the accumulated operating time of the air conditioner at the first moment is smaller, and the accumulated operating time of the air conditioner at the second moment is larger.
[0132] Therefore, whether the target operation stage includes a dynamic stage and a steady-state stage can be determined based on the size relationship between the accumulated operation time of the air conditioner and the set time.
[0133] For example, let's set the duration to 30 minutes.
[0134] Continuing with the operation data set A as an example, the cumulative operating time of the air conditioner at sampling times 1 to 4 is less than or equal to 30 minutes, and the cumulative operating time of the air conditioner at sampling times 5 to 7 is greater than 30 minutes. Then sampling times 1 to 4 are all the first moments, and sampling times 5 to 7 are all the second moments. It is determined that the target operating stage of the air conditioner corresponding to the operation data set A includes a dynamic stage and a steady-state stage.
[0135] Continuing with the operation data set B as an example, the cumulative operation time of the air conditioner at sampling times 1 to 4 is less than or equal to 30 minutes, and the sampling times 1 to 4 are all the first moments, and it is determined that the target operation stage of the air conditioner corresponding to the operation data set B includes the dynamic stage.
[0136] Continuing with the operation data set C as an example, the cumulative operation time of the air conditioner at sampling times 5 to 7 is greater than 30 minutes, and the sampling times 5 to 7 are all the second moments, and it is determined that the target operation stage of the air conditioner corresponding to the operation data set C includes the steady-state stage.
[0137] Method 2: If the operating data set includes operating data at a third moment, wherein multiple sampling moments include the third moment, and the absolute value of the difference between the indoor ambient temperature and the set temperature at the third moment is greater than the set threshold, it is determined that the target operating stage includes a dynamic stage; if the operating data set includes operating data at a fourth moment, wherein multiple sampling moments include the fourth moment, and the absolute value of the difference at the fourth moment is less than or equal to the set threshold, it is determined that the target operating stage includes a steady-state stage.
[0138] It can be understood that the difference between the indoor ambient temperature and the set temperature at the third moment is relatively large, and the difference between the indoor ambient temperature and the set temperature at the fourth moment is relatively small.
[0139] Thus, it is possible to determine whether the target operation phase includes a dynamic phase and a steady-state phase, taking into account the deviation between the indoor ambient temperature and the set temperature.
[0140] For example, the threshold is set to 0.5°C and the temperature is set to 22°C.
[0141] Continuing with the operation data set A as an example, the indoor ambient temperatures at sampling times 1 to 7 are 30°C, 25°C, 24°C, 23°C, 22.4°C, 22°C, and 22.2°C, respectively. The absolute value of the difference between the indoor ambient temperature and 22°C at sampling times 1 to 4 is greater than 0.5°C, and the absolute value of the difference between the indoor ambient temperature and 22°C at sampling times 5 to 7 is less than or equal to 0.5°C. Sampling times 1 to 4 are all the third moments, and sampling times 5 to 7 are all the fourth moments. It is determined that the target operation stage of the air conditioner corresponding to the operation data set A includes a dynamic stage and a steady-state stage.
[0142] Continuing with the operation data set B as an example, the indoor ambient temperatures at sampling times 1 to 4 are 30°C, 25°C, 24°C, and 23°C, respectively. The absolute value of the difference between the indoor ambient temperature at sampling times 1 to 4 and 22°C is greater than 0.5°C, and the sampling times 1 to 4 are all the third moments. It is determined that the target operation stage of the air conditioner corresponding to the operation data set B includes the dynamic stage.
[0143] Continuing with the operation data set C as an example, the indoor ambient temperatures at sampling times 5 to 7 are 22.4°C, 22°C, and 22.2°C, respectively. The absolute value of the difference between the indoor ambient temperature and 22°C at sampling times 5 to 7 is less than or equal to 0.5°C. Then, sampling times 5 to 7 are all the fourth moments, and it is determined that the target operation stage of the air conditioner corresponding to the operation data set C includes a steady-state stage.
[0144] Figure 2 is a flow chart showing a method for diagnosing a fault of an air conditioner according to another exemplary embodiment. Figure 2 As shown, the fault diagnosis method of the air conditioner according to the embodiment of the present disclosure includes the following steps.
[0145] S201, obtaining an operation data set corresponding to an air conditioner, wherein the operation data set includes operation data of the air conditioner at multiple sampling moments.
[0146] For the relevant contents of step S201, please refer to the above embodiment, which will not be repeated here.
[0147] S202, based on the target operation stage, determining the target feature category to be extracted from the operation data set as a target feature extraction method.
[0148] It should be noted that the number of target feature categories is at least one, and there is no excessive limitation on the target feature categories. For example, the target feature categories may include at least one of dynamic feature categories, steady-state feature categories, and cumulative feature categories. For example, the dynamic feature category includes the change speed and / or change acceleration of the operating data of the air conditioner. For the relevant contents of the dynamic feature category, the steady-state feature category, and the cumulative feature category, please refer to the above embodiments, and will not be repeated here.
[0149] In one embodiment, based on the target operation stage, determining the target feature category to be extracted from the operation data set includes obtaining a mapping relationship between the candidate operation stages of the air conditioner and the candidate feature categories, and obtaining the candidate feature category having a mapping relationship with the target operation stage as the target feature category. It is understandable that different candidate operation stages may be mapped to different candidate feature categories.
[0150] In one embodiment, based on the target operation stage, the target feature category to be extracted from the operation data set is determined, including if the target operation stage includes a dynamic stage, determining that the target feature category includes a dynamic feature category and a cumulative feature category, and if the target operation stage includes a steady-state stage, determining that the target feature category includes a steady-state feature category and a cumulative feature category. Thus, when the target operation stage includes a dynamic stage, determining that the target feature category includes a dynamic feature category and a cumulative feature category, and when the target operation stage includes a steady-state stage, determining that the target feature category includes a steady-state feature category and a cumulative feature category.
[0151] For example, continuing to take the operation data set B as an example, the target operation stage of the air conditioner corresponding to the operation data set B includes a dynamic stage, and it can be determined that the target feature category includes a dynamic feature category and a cumulative feature category.
[0152] For example, continuing to take the operation data set C as an example, the target operation stage of the air conditioner corresponding to the operation data set C includes a steady-state stage, and it can be determined that the target feature category includes a steady-state feature category and a cumulative feature category.
[0153] It is understandable that the heat conduction process of the air conditioner is in a constant state of change during the dynamic stage. And because the parameters of each air conditioner and room are different, the corresponding specific heat capacity C and thermal resistance R are different, and the statistical values of the operating data (such as the value, average, variance, mode, maximum, minimum, etc. of the operating data) have many changes and cannot directly reflect whether the air conditioner is faulty.
[0154] However, the statistical value of the operating data combined with the change speed and / or change acceleration of each operating data of the air conditioner at this time can reflect whether the air conditioner has a fault. For example, when the compressor frequency freq is relatively large, the change speed of the inner tube temperature at this time The absolute value should be relatively large to be reasonable, otherwise it means that the air conditioner may have a fluorine deficiency failure.
[0155] Therefore, the dynamic feature category at the sampling moment may include the value of the operating data at the sampling moment, as well as the average value, variance, mode, maximum value, minimum value, etc. of multiple operating data in the time period of the sampling moment, and the change speed and / or change acceleration of the operating data at the sampling moment.
[0156] In one embodiment, the dynamic feature is obtained by extracting at least one operating data of indoor ambient temperature, outdoor ambient temperature, compressor frequency, exhaust temperature, inner pipe temperature, outdoor unit energy consumption power, and indoor fan speed.
[0157] For example, the operating data is the indoor ambient temperature T iTaking (t) as an example, the dynamic characteristics of the indoor ambient temperature at sampling time t2 include [d1, d2, d3, d4, d5, d6, d7].
[0158] d1=T i (t2)
[0159]
[0160] d3=max(T i (t2-5)...T i (t2+5))
[0161] d4=min(T i (t2-5)...T i (t2+5))
[0162] d5=mode(T i (t2-5)...T i (t2+5))
[0163]
[0164] Among them, T i (t2) is the indoor ambient temperature at sampling time t2, T i (t2-m) is the indoor ambient temperature m seconds before sampling time t2, T i (t2-n) is the indoor ambient temperature n seconds after sampling time t2, m and n are both positive integers not greater than 5, max(·) is the maximum value operator, min(·) is the minimum value operator, mode(·) is the mode operator, in this embodiment, T=1, d6 is the changing speed of the indoor ambient temperature at sampling time t2, and d7 is the changing acceleration of the indoor ambient temperature at sampling time t2.
[0165] It is understandable that the heat conduction process of the air conditioner is in a stable state in the steady-state stage, at which time the compressor operating frequency, inner tube temperature, indoor ambient temperature, exhaust temperature, outdoor unit energy consumption power, etc. remain stable, and the statistical value of the operating data in the steady-state stage (such as the value, average value, variance, mode, maximum value, minimum value, etc. of the operating data) can reflect whether the air conditioner is faulty. Therefore, the steady-state feature category at the sampling time may include the value of the operating data at the sampling time, and the average value, variance, mode, maximum value, minimum value, etc. of multiple operating data in the time period of the sampling time.
[0166] In one embodiment, the steady-state feature is obtained by extracting at least one operating data of indoor ambient temperature, outdoor ambient temperature, compressor frequency, exhaust temperature, inner pipe temperature, outdoor unit energy consumption power, and indoor fan speed.
[0167] For example, the operating data is the indoor ambient temperature T i (t) as an example, the steady-state characteristics of the indoor ambient temperature at sampling time t4 include [x1, x2, x3, x4, x5].
[0168] x1=T i (t4)
[0169]
[0170] x3=max(T i (t4-5)...T i (t4+5))
[0171] x4=min(T i (t4-5)...T i (t4+5))
[0172] x5=mode(T i (t4-5)...T i (t4+5))
[0173] Among them, T i (t4) are the indoor ambient temperatures at sampling time t4, T i (t4-m) is the indoor ambient temperature m seconds before sampling time t4, T i (t4-n) is the indoor ambient temperature n seconds after sampling time t4, m and n are both positive integers not greater than 5, max(·) is the maximum value operator, min(·) is the minimum value operator, and mode(·) is the mode operator.
[0174] It is understandable that the cumulative feature mainly solves the problem of inaccurate steady-state feature sampling due to compressor frequency fluctuations in some control scenarios, which causes the air conditioner to be in an unstable state. Cumulative features need to be extracted in both the dynamic stage and the steady-state stage. The cumulative feature category at the sampling time may include the cumulative change in the operating data from the time the air conditioner is turned on to the sampling time.
[0175] In one embodiment, the cumulative feature is obtained by extracting at least one of the operating data of indoor ambient temperature, outdoor ambient temperature, compressor frequency, inner tube temperature, and outdoor machine energy consumption power, that is, the cumulative feature category at the sampling time may include the cumulative changes of compressor frequency, outdoor machine energy consumption power, inner tube temperature, indoor ambient temperature, and outdoor ambient temperature from the start-up time of the air conditioner to the sampling time. Among them, the cumulative changes of compressor frequency, outdoor machine energy consumption power, and outdoor ambient temperature can all reflect the power consumption of the air conditioner, and the cumulative changes of inner tube temperature and indoor ambient temperature can all reflect the actual cooling capacity or heating capacity output by the air conditioner.
[0176] For example, the accumulated features at sampling time t2 include [a1, a2, a3, a4, a5].
[0177]
[0178] Among them, a1 is the cumulative change of the compressor frequency from the start-up time t=0 of the air conditioner to the sampling time t2, a2 is the cumulative change of the outdoor unit energy consumption power from the start-up time of the air conditioner to the sampling time t2, a3 is the cumulative change of the inner pipe temperature from the start-up time of the air conditioner to the sampling time t2, a4 is the cumulative change of the indoor ambient temperature from the start-up time of the air conditioner to the sampling time t2, and a5 is the cumulative change of the outdoor ambient temperature from the start-up time of the air conditioner to the sampling time t2.
[0179] S203, extracting features from the operation data set to obtain operation features of the target feature category.
[0180] S204, adding the operation features of the target feature category to the operation feature set.
[0181] For example, continuing to take the operation data set B as an example, the target operation stage of the air conditioner corresponding to the operation data set B includes a dynamic stage, and it can be determined that the target feature category includes a dynamic feature category and a cumulative feature category. Feature extraction can be performed on the operation data set B to obtain dynamic features and cumulative features, and the dynamic features and cumulative features are added to the operation feature set E. Among them, the operation feature set E may include dynamic features at sampling times 1 to 4, and cumulative features at sampling times 1 to 4.
[0182] Among them, the dynamic characteristics at the sampling time include the dynamic characteristics of the indoor ambient temperature, outdoor ambient temperature, compressor frequency, exhaust temperature, inner pipe temperature, outdoor unit energy consumption power, and indoor fan speed at the sampling time, and the cumulative characteristics at the sampling time include the cumulative characteristics of the indoor ambient temperature, outdoor ambient temperature, compressor frequency, inner pipe temperature, and outdoor unit energy consumption power at the sampling time.
[0183] For example, continuing to take the operation data set C as an example, the target operation stage of the air conditioner corresponding to the operation data set C includes a steady-state stage, and it can be determined that the target feature category includes a steady-state feature category and a cumulative feature category. Feature extraction can be performed on the operation data set C to obtain steady-state features and cumulative features, and the steady-state features and cumulative features are added to the operation feature set F. Among them, the operation feature set F may include steady-state features at sampling times 5 to 7, and cumulative features at sampling times 5 to 7.
[0184] Among them, the steady-state characteristics at the sampling time include the steady-state characteristics of the indoor ambient temperature, outdoor ambient temperature, compressor frequency, exhaust temperature, inner pipe temperature, outdoor unit energy consumption power, and indoor fan speed at the sampling time, and the cumulative characteristics at the sampling time include the cumulative characteristics of the indoor ambient temperature, outdoor ambient temperature, compressor frequency, inner pipe temperature, and outdoor unit energy consumption power at the sampling time.
[0185] S205: Perform fault diagnosis on the air conditioner based on the operation feature set.
[0186] For the relevant contents of step S205, please refer to the above embodiment, which will not be repeated here.
[0187] The fault diagnosis method for an air conditioner provided by an embodiment of the present disclosure determines the target feature category to be extracted from the operating data set based on the target operating stage, performs feature extraction on the operating data set as a target feature extraction method, obtains the operating features of the target feature category, and adds the operating features of the target feature category to the operating feature set. Thus, the target operating stage of the air conditioner corresponding to the operating data set can be taken into consideration to determine the target feature category to be extracted from the operating data set, so as to perform feature extraction on the operating data set, and accurately obtain the target feature category to be extracted, which helps to collect a small number of features that are strongly related to the fault diagnosis of the air conditioner.
[0188] Figure 3 is a flow chart showing a method for diagnosing a fault of an air conditioner according to another exemplary embodiment. Figure 3 As shown, the fault diagnosis method of the air conditioner according to the embodiment of the present disclosure includes the following steps.
[0189] S301, obtaining an operation data set corresponding to an air conditioner, wherein the operation data set includes operation data of the air conditioner at multiple sampling moments.
[0190] S302: If there are multiple target operation stages, split the operation data set to obtain multiple operation sub-data sets, wherein the operation sub-data sets correspond to the target operation stages one by one.
[0191] For the relevant contents of steps S301 - S302 , please refer to the above embodiments and will not be described again here.
[0192] S303: Determine a sub-feature category to be extracted from the operation sub-dataset based on the target operation stage corresponding to the operation sub-dataset.
[0193] S304, extracting features from the operation sub-dataset to obtain operation features of the sub-feature category.
[0194] It should be noted that the target feature category includes sub-feature categories, and different operating sub-data sets may correspond to different sub-feature categories. The relevant content of step S303 can refer to the relevant content of step S202, which will not be repeated here.
[0195] For example, continuing to take the operation data set A as an example, the target operation stage of the air conditioner corresponding to the operation data set A includes the dynamic stage and the steady-state stage. The operation data set A can be split to obtain operation sub-datasets A1 and A2. The operation sub-dataset A1 includes the operation data of the air conditioner at sampling times 1 to 4, and the operation sub-dataset A2 includes the operation data of the air conditioner at sampling times 5 to 7. The target operation stage corresponding to the operation sub-dataset A1 is the dynamic stage, and the target operation stage corresponding to the operation sub-dataset A2 is the steady-state stage.
[0196] Based on the target operation stage (i.e., dynamic stage) corresponding to the operation sub-dataset A1, it can be determined that the sub-feature categories to be extracted from the operation sub-dataset A1 include dynamic feature categories and cumulative feature categories. Feature extraction is performed on the operation sub-dataset A1 to obtain dynamic features and cumulative features, and the dynamic features and cumulative features are added to the operation feature set D1. The operation feature set D1 may include dynamic features at sampling times 1 to 4, and cumulative features at sampling times 1 to 4.
[0197] Among them, the dynamic characteristics at the sampling time include the dynamic characteristics of the indoor ambient temperature, outdoor ambient temperature, compressor frequency, exhaust temperature, inner pipe temperature, outdoor unit energy consumption power, and indoor fan speed at the sampling time, and the cumulative characteristics at the sampling time include the cumulative characteristics of the indoor ambient temperature, outdoor ambient temperature, compressor frequency, inner pipe temperature, and outdoor unit energy consumption power at the sampling time.
[0198] Based on the target operation stage (i.e., steady-state stage) corresponding to the operation sub-dataset A2, it can be determined that the sub-feature categories to be extracted from the operation sub-dataset A2 include steady-state feature categories and cumulative feature categories. Feature extraction can be performed on the operation sub-dataset A2 to obtain steady-state features and cumulative features, and the steady-state features and cumulative features are added to the operation feature set D2. The operation feature set D2 may include steady-state features at sampling times 5 to 7, and cumulative features at sampling times 5 to 7.
[0199] Among them, the steady-state characteristics at the sampling time include the steady-state characteristics of the indoor ambient temperature, outdoor ambient temperature, compressor frequency, exhaust temperature, inner pipe temperature, outdoor unit energy consumption power, and indoor fan speed at the sampling time, and the cumulative characteristics at the sampling time include the cumulative characteristics of the indoor ambient temperature, outdoor ambient temperature, compressor frequency, inner pipe temperature, and outdoor unit energy consumption power at the sampling time.
[0200] Continuing to take the running data set A as an example, the feature extraction method of the running data set A is shown in Table 1.
[0201] Table 1 Feature extraction method for running dataset A
[0202]
[0203] As shown in Table 1, dynamic features at sampling times 1 to 4, steady-state features at sampling times 5 to 7, and cumulative features at sampling times 1 to 7 can be extracted from the operation data set A. The dynamic features at the sampling time include the dynamic features of the indoor ambient temperature, outdoor ambient temperature, compressor frequency, exhaust temperature, inner pipe temperature, outdoor unit energy consumption power, and indoor fan speed at the sampling time, the steady-state features at the sampling time include the steady-state features of the indoor ambient temperature, outdoor ambient temperature, compressor frequency, exhaust temperature, inner pipe temperature, outdoor unit energy consumption power, and indoor fan speed at the sampling time, and the cumulative features at the sampling time include the cumulative features of the indoor ambient temperature, outdoor ambient temperature, compressor frequency, inner pipe temperature, and outdoor unit energy consumption power at the sampling time.
[0204] S305: Add the operation features of the target feature category to the operation feature set.
[0205] S306: Perform fault diagnosis on the air conditioner based on the operation feature set.
[0206] For the relevant contents of steps S305-S306, please refer to the above embodiment and will not be repeated here.
[0207] The fault diagnosis method for an air conditioner provided by an embodiment of the present disclosure, if there are multiple target operation stages, splits the operation data set to obtain multiple operation sub-data sets, wherein the operation sub-data sets correspond one-to-one to the target operation stages, and based on the target operation stages corresponding to the operation sub-data sets, determines the sub-feature categories to be extracted from the operation sub-data sets, performs feature extraction on the operation sub-data sets, and obtains the operation features of the sub-feature categories. Thus, when there are multiple target operation stages, the operation data set can be split according to the target operation stages to obtain multiple operation sub-data sets, and considering the target operation stages corresponding to the operation sub-data sets, determines the sub-feature categories to be extracted from the operation sub-data sets, and performs feature extraction on each operation sub-data set respectively, so that the sub-feature categories to be extracted can be accurately obtained, which helps to collect a small number of features that are strongly related to the fault diagnosis of the air conditioner.
[0208] Figure 4 is a flow chart of a model training method according to an exemplary embodiment. Figure 4 As shown, the model training method of the embodiment of the present disclosure includes the following steps.
[0209] S401, obtaining a sample operation data set corresponding to a sample air conditioner, wherein the sample operation data set includes sample operation data of the sample air conditioner at multiple sampling moments.
[0210] S402, determining a sample feature extraction method of the sample operation data set based on operations of the sample user on the sample client corresponding to the sample air conditioner and / or based on the sample operation stage of the sample air conditioner corresponding to the sample operation data set.
[0211] S403, extracting features from the sample operation data set according to the sample feature extraction method to obtain a sample operation feature set corresponding to the sample air conditioner, wherein the sample operation feature set includes sample operation features of the sample air conditioner at multiple sampling moments.
[0212] It should be noted that the execution subject of the model training method of the embodiment of the present disclosure is an electronic device, such as a mobile phone, a notebook, a desktop computer, a vehicle-mounted terminal, a smart home appliance, a wearable device, etc. Among them, the wearable device may include a wrist-worn device (such as a smart watch, a smart bracelet), a head-worn device, a foot-worn device, etc. The model training method of the embodiment of the present disclosure can be executed by the model training device of the embodiment of the present disclosure, and the model training device of the embodiment of the present disclosure can be configured in any electronic device to execute the model training method of the embodiment of the present disclosure.
[0213] It should be noted that, for the relevant content of step S401, refer to the relevant content of step S101 in the above embodiment, for the relevant content of step S402, refer to the relevant content of step S102 in the above embodiment, and for the relevant content of step S403, refer to the relevant content of step S103 in the above embodiment, which will not be repeated here.
[0214] In one embodiment, based on the operation of the sample user on the sample client corresponding to the sample air conditioner, and / or based on the sample operation stage of the sample air conditioner corresponding to the sample operation data set, the sample feature extraction method of the sample operation data set is determined, including controlling the display of multiple first candidate selection items on the interactive page of the sample client, wherein the first candidate selection item is used to characterize the candidate feature extraction method, and based on the selection operation of the sample user on the interactive page for the first target selection item, the candidate feature extraction method corresponding to the first target selection item is used as the sample feature extraction method. Thus, the candidate feature extraction method selected by the sample user can be used as the sample feature extraction method, and personalized acquisition of the sample feature extraction method can be achieved.
[0215] In one embodiment, based on the operations of the sample user on the sample client corresponding to the sample air conditioner, and / or based on the sample operation stage of the sample air conditioner corresponding to the sample operation data set, a sample feature extraction method of the sample operation data set is determined, including determining the sample feature category to be extracted from the sample operation data set based on the sample operation stage as a sample feature extraction method.
[0216] According to the sample feature extraction method, feature extraction is performed on the sample operation data set to obtain a sample operation feature set corresponding to the sample air conditioner, including feature extraction on the sample operation data set to obtain sample operation features of the sample feature category, and the sample operation features of the sample feature category are added to the sample operation feature set.
[0217] Therefore, the sample operation stage of the air conditioner corresponding to the sample operation data set can be taken into consideration to determine the sample feature category to be extracted from the sample operation data set, so as to perform feature extraction on the sample operation data set. The sample feature category to be extracted can be accurately obtained, which is helpful to collect a small number of features that are strongly related to the fault diagnosis of the air conditioner.
[0218] In some examples, based on the sample operation stage, the sample feature category to be extracted from the sample operation data set is determined, including if there are multiple sample operation stages, splitting the sample operation data set to obtain multiple sample operation sub-data sets, wherein the sample operation sub-data sets correspond one-to-one to the sample operation stage, and obtaining the sample feature category determined by the sample operation stage corresponding to the sample operation sub-data set as the sample sub-feature category to be extracted from the sample operation sub-data set.
[0219] Feature extraction is performed on the sample running data set to obtain sample running features of the sample feature category, including feature extraction on the sample running sub-data set to obtain sample running features of the sample sub-feature category.
[0220] Therefore, when there are multiple sample operation stages, the sample operation data set can be split according to the sample operation stages to obtain multiple sample operation sub-data sets, and the sample sub-feature categories to be extracted from the sample operation sub-data sets are determined taking into account the sample operation stages corresponding to the sample operation sub-data sets, so as to perform feature extraction on each sample operation sub-data set respectively, and the sample sub-feature categories to be extracted can be accurately obtained, which is helpful to collect a small number of features that are strongly related to the fault diagnosis of the air conditioner.
[0221] S404: training a fault diagnosis model based on the sample running feature set.
[0222] It should be noted that the fault diagnosis model can be trained based on the sample operation feature set by using any model training method in the relevant technology, and no excessive restrictions are made here.
[0223] In one embodiment, the sample operation data set also includes sample fault diagnosis results of the sample air conditioner. Based on the sample operation feature set, a fault diagnosis model is trained, including inputting the sample operation feature set into the fault diagnosis model, and having the fault diagnosis model output predicted fault diagnosis results of the sample air conditioner. The fault diagnosis model is trained based on the predicted fault diagnosis results of the sample air conditioner and the sample fault diagnosis results of the sample air conditioner.
[0224] In some examples, the fault diagnosis model is trained based on the predicted fault diagnosis results of the sample air conditioner and the sample fault diagnosis results of the sample air conditioner, including obtaining a loss function of the fault diagnosis model based on the predicted fault diagnosis results of the sample air conditioner and the sample fault diagnosis results of the sample air conditioner, and training the fault diagnosis model based on the loss function. It should be noted that there are no excessive restrictions on the loss function, for example, it can include CE (Cross Entropy), MSE (Mean-Square Error), KL (Kullback-Leibler) divergence, contrast loss function, etc. Among them, CE can include multi-classification cross entropy.
[0225] The model training method provided by the embodiment of the present disclosure obtains a sample operation data set corresponding to a sample air conditioner, wherein the sample operation data set includes sample operation data of the sample air conditioner at multiple sampling moments, determines a sample feature extraction method of the sample operation data set based on the operation of the sample user on the sample client corresponding to the sample air conditioner, and / or based on the sample operation stage of the sample air conditioner corresponding to the sample operation data set, performs feature extraction on the sample operation data set according to the sample feature extraction method, obtains a sample operation feature set corresponding to the sample air conditioner, wherein the sample operation feature set includes sample operation features of the sample air conditioner at multiple sampling moments, and trains a fault diagnosis model based on the sample operation feature set. Thus, the sample feature extraction method can be determined considering the operation of the sample user on the interactive page and / or the sample operation stage of the air conditioner corresponding to the sample operation data set, so as to extract features from the sample operation data set, thereby improving the accuracy of the sample feature extraction method, facilitating the collection of sample features that are relatively small in number and strongly related to the fault diagnosis of the air conditioner, and using the sample operation feature set to train the fault diagnosis model, thereby improving the training accuracy and efficiency of the fault diagnosis model and saving computing resources.
[0226] Figure 5 FIG. 1 is a block diagram of a fault diagnosis device for an air conditioner according to an exemplary embodiment. Figure 5 The fault diagnosis device 100 of the air conditioner according to the embodiment of the present disclosure includes: an acquisition module 110 , a determination module 120 , a feature extraction module 130 and a fault diagnosis module 140 .
[0227] The acquisition module 110 is configured to acquire an operation data set corresponding to the air conditioner, wherein the operation data set includes operation data of the air conditioner at multiple sampling moments;
[0228] The determination module 120 is configured to perform a target feature extraction method of the operation data set based on the user's operation on the target client corresponding to the air conditioner and / or based on the target operation stage of the air conditioner corresponding to the operation data set;
[0229] The feature extraction module 130 is configured to perform feature extraction on the operation data set according to the target feature extraction method to obtain an operation feature set corresponding to the air conditioner, wherein the operation feature set includes operation features of the air conditioner at multiple sampling moments;
[0230] The fault diagnosis module 140 is configured to perform fault diagnosis on the air conditioner based on the operation feature set.
[0231] In one embodiment of the present disclosure, the determination module 120 is further configured to execute: controlling the display of multiple first candidate selection items on the interactive page of the target client, wherein the first candidate selection items are used to represent candidate feature extraction methods; based on the user's selection operation on the interactive page for the first target selection item, using the candidate feature extraction method corresponding to the first target selection item as the target feature extraction method.
[0232] In one embodiment of the present disclosure, the acquisition module 110 is further configured to perform: controlling the display of multiple second candidate selection items and multiple third candidate selection items on the interactive page of the target client, wherein the second candidate selection items are used to represent candidate sampling time periods, and the third candidate selection items are used to represent candidate categories of the operating data of the air conditioner; based on the user's selection operation on the second target selection item on the interactive page, the candidate sampling time period corresponding to the second target selection item is used as the target sampling time period; based on the user's selection operation on the third target selection item on the interactive page, the candidate category corresponding to the third target selection item is used as the target category; and adding the operating data of the target category of the air conditioner under the target sampling time period to the operating data set.
[0233] In one embodiment of the present disclosure, before acquiring the operating data set corresponding to the air conditioner, the acquisition module 110 is further configured to execute: detecting that the user has activated the fault diagnosis function on the interactive page of the target client.
[0234] In one embodiment of the present disclosure, the determination module 120 is further configured to perform: based on the target operation stage, determining the target feature category to be extracted from the operation data set as the target feature extraction method;
[0235] The feature extraction module 130 is further configured to execute: extracting features from the operation data set to obtain the operation features of the target feature category; and adding the operation features of the target feature category to the operation feature set.
[0236] In one embodiment of the present disclosure, the determination module 120 is further configured to execute: if there are multiple target operation stages, split the operation data set to obtain multiple operation sub-data sets, wherein the operation sub-data sets correspond one-to-one to the target operation stages;
[0237] The feature extraction module 130 is further configured to execute: obtaining a target feature category determined by a target operation stage corresponding to the operation sub-dataset as a sub-feature category to be extracted from the operation sub-dataset; and performing feature extraction on the operation sub-dataset to obtain operation features of the sub-feature category.
[0238] In one embodiment of the present disclosure, the determination module 120 is further configured to execute: obtaining a mapping relationship between candidate operating stages of the air conditioner and candidate feature categories; obtaining a candidate feature category having a mapping relationship with the target operating stage as the target feature category.
[0239] In one embodiment of the present disclosure, the target feature category includes at least one of a dynamic feature category, a steady-state feature category, and a cumulative feature category.
[0240] In one embodiment of the present disclosure, the determination module 120 is further configured to execute: if the target operation stage includes a dynamic stage, determining that the target feature category includes the dynamic feature category and the cumulative feature category; if the target operation stage includes a steady-state stage, determining that the target feature category includes the steady-state feature category and the cumulative feature category.
[0241] In one embodiment of the present disclosure, the dynamic feature category includes a change speed and / or a change acceleration of the operating data of the air conditioner.
[0242] In one embodiment of the present disclosure, the determination module 120 is further configured to execute: if the operating data set includes operating data at a first moment, wherein multiple sampling moments include the first moment, and the cumulative operating time of the air conditioner at the first moment is less than or equal to the set time, determining that the target operating stage includes a dynamic stage; if the operating data set includes operating data at a second moment, wherein multiple sampling moments include the second moment, and the cumulative operating time of the air conditioner at the second moment is greater than the set time, determining that the target operating stage includes a steady-state stage.
[0243] In one embodiment of the present disclosure, the determination module 120 is further configured to execute: if the operating data set includes operating data at a third moment, wherein multiple sampling moments include the third moment, and the absolute value of the difference between the indoor ambient temperature and the set temperature at the third moment is greater than the set threshold, it is determined that the target operating stage includes a dynamic stage; if the operating data set includes operating data at a fourth moment, wherein multiple sampling moments include the fourth moment, and the absolute value of the difference at the fourth moment is less than or equal to the set threshold, it is determined that the target operating stage includes a steady-state stage.
[0244] In one embodiment of the present disclosure, the fault diagnosis module 140 is further configured to execute: inputting the operating feature set into a target fault diagnosis model, and having the target fault diagnosis model output a fault diagnosis result of the air conditioner.
[0245] In one embodiment of the present disclosure, the fault diagnosis module 140 is further configured to execute: controlling the display of multiple fourth candidate selection items on the interactive page of the target client, wherein the fourth candidate selection items are used to characterize candidate fault diagnosis models; based on the user's selection operation on the fourth target selection item on the interactive page, using the candidate fault diagnosis model corresponding to the fourth target selection item as the target fault diagnosis model.
[0246] In one embodiment of the present disclosure, the fault diagnosis module 140 is further configured to execute: controlling the display of fault diagnosis data of the air conditioner on the interactive page of the target client, wherein the fault diagnosis data includes the target feature extraction method and / or the fault diagnosis result of the air conditioner.
[0247] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.
[0248] The fault diagnosis device for an air conditioner provided by the embodiment of the present disclosure obtains an operation data set corresponding to the air conditioner, wherein the operation data set includes the operation data of the air conditioner at multiple sampling moments, determines a target feature extraction method for the operation data set based on the user's operation on the target client corresponding to the air conditioner, and / or based on the target operation stage of the air conditioner corresponding to the operation data set, performs feature extraction on the operation data set according to the target feature extraction method, obtains an operation feature set corresponding to the air conditioner, wherein the operation feature set includes the operation features of the air conditioner at multiple sampling moments, and performs fault diagnosis on the air conditioner based on the operation feature set. Thus, the target feature extraction method can be determined to extract features from the operation data set in consideration of the user's operation on the interactive page and / or the target operation stage of the air conditioner corresponding to the operation data set, thereby improving the accuracy of the target feature extraction method, facilitating the collection of features that are relatively small in number and strongly related to the fault diagnosis of the air conditioner, and automatically performing fault diagnosis on the air conditioner using the operation feature set, thereby improving the accuracy and efficiency of the fault diagnosis of the air conditioner and saving computing resources.
[0249] Figure 6 is a block diagram of a model training device according to an exemplary embodiment. Figure 6 The model training device 200 of the embodiment of the present disclosure includes: an acquisition module 210, a determination module 220, a feature extraction module 230 and a training module 240.
[0250] The acquisition module 210 is configured to acquire a sample operation data set corresponding to the sample air conditioner, wherein the sample operation data set includes sample operation data of the sample air conditioner at multiple sampling moments;
[0251] The determination module 220 is configured to perform operations performed by the sample user on the sample client corresponding to the sample air conditioner and / or determine a sample feature extraction method of the sample operation data set based on the sample operation stage of the sample air conditioner corresponding to the sample operation data set;
[0252] The feature extraction module 230 is configured to perform feature extraction on the sample operation data set according to the sample feature extraction method to obtain a sample operation feature set corresponding to the sample air conditioner, wherein the sample operation feature set includes sample operation features of the sample air conditioner at multiple sampling moments;
[0253] The training module 240 is configured to execute training of the fault diagnosis model based on the sample running feature set.
[0254] In one embodiment of the present disclosure, the determination module 220 is further configured to execute: controlling the display of multiple first candidate selection items on the interactive page of the sample client, wherein the first candidate selection items are used to characterize candidate feature extraction methods; based on the selection operation of the sample user on the interactive page for the first target selection item, using the candidate feature extraction method corresponding to the first target selection item as the sample feature extraction method.
[0255] In one embodiment of the present disclosure, the determination module 220 is further configured to execute: based on the sample operation stage, determining the sample feature category to be extracted from the sample operation data set as the sample feature extraction method;
[0256] The feature extraction module 230 is further configured to execute: extracting features from the sample operation data set to obtain sample operation features of the sample feature category; and adding the sample operation features of the sample feature category to the sample operation feature set.
[0257] In one embodiment of the present disclosure, the sample operation data set also includes sample fault diagnosis results of the sample air conditioner, and the training module 240 is further configured to perform: inputting the sample operation feature set into the fault diagnosis model, and the fault diagnosis model outputs the predicted fault diagnosis results of the sample air conditioner; and training the fault diagnosis model based on the predicted fault diagnosis results of the sample air conditioner and the sample fault diagnosis results of the sample air conditioner.
[0258] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.
[0259] The model training device provided by the embodiment of the present disclosure obtains a sample operation data set corresponding to a sample air conditioner, wherein the sample operation data set includes sample operation data of the sample air conditioner at multiple sampling moments, determines a sample feature extraction method of the sample operation data set based on the operation of the sample user on the sample client corresponding to the sample air conditioner, and / or based on the sample operation stage of the sample air conditioner corresponding to the sample operation data set, performs feature extraction on the sample operation data set according to the sample feature extraction method, obtains a sample operation feature set corresponding to the sample air conditioner, wherein the sample operation feature set includes sample operation features of the sample air conditioner at multiple sampling moments, and trains a fault diagnosis model based on the sample operation feature set. Thus, the sample feature extraction method can be determined considering the operation of the sample user on the interactive page and / or the sample operation stage of the air conditioner corresponding to the sample operation data set, so as to extract features from the sample operation data set, thereby improving the accuracy of the sample feature extraction method, facilitating the collection of sample features that are relatively small in number and strongly related to the fault diagnosis of the air conditioner, and using the sample operation feature set to train the fault diagnosis model, thereby improving the training accuracy and efficiency of the fault diagnosis model and saving computing resources.
[0260] Figure 7 It is a block diagram of an electronic device according to an exemplary embodiment.
[0261] like Figure 7 As shown, the electronic device 300 includes:
[0262] The memory 310 and the processor 320, a bus 330 connecting different components (including the memory 310 and the processor 320), the memory 310 stores a computer program, and when the processor 320 executes the program, the fault diagnosis method and model training method of the air conditioner described in the embodiment of the present disclosure are implemented.
[0263] Bus 330 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. For example, these architectures include but are not limited to Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus and Peripheral Component Interconnect (PCI) bus.
[0264] The electronic device 300 typically includes a variety of electronic device readable media. These media can be any available media that can be accessed by the electronic device 300, including volatile and non-volatile media, removable and non-removable media.
[0265] The memory 310 may also include computer system readable media in the form of volatile memory, such as random access memory (RAM) 340 and / or cache memory 350. The electronic device 300 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 360 may be used to read and write non-removable, non-volatile magnetic media ( Figure 7 not shown, usually called a "hard drive"). Although Figure 7 Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 330 via one or more data medium interfaces. The memory 310 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present disclosure.
[0266] A program / utility 380 having a set (at least one) of program modules 370 may be stored, for example, in the memory 310, such program modules 370 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. The program modules 370 generally perform the functions and / or methods of the embodiments described in the present disclosure.
[0267] The electronic device 300 may also communicate with one or more external devices 390 (e.g., keyboard, pointing device, display 391, etc.), one or more devices that enable a user to interact with the electronic device 300, and / or any device that enables the electronic device 300 to communicate with one or more other computing devices (e.g., network card, modem, etc.). Such communication may be performed through an input / output (I / O) interface 392. Furthermore, the electronic device 300 may also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN) and / or public network, such as the Internet) through a network adapter 393. Figure 7 As shown, the network adapter 393 communicates with other modules of the electronic device 300 via the bus 330. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with the electronic device 300, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.
[0268] The processor 320 executes various functional applications and data processing by running the programs stored in the memory 310 .
[0269] It should be noted that the implementation process and technical principles of the electronic device of this embodiment refer to the aforementioned explanation of the fault diagnosis method and model training method of the air conditioner in the embodiment of the present disclosure, and will not be repeated here.
[0270] The electronic device provided by the embodiment of the present disclosure can execute the fault diagnosis method of the air conditioner as described above, obtain the operation data set corresponding to the air conditioner, wherein the operation data set includes the operation data of the air conditioner at multiple sampling moments, determine the target feature extraction method of the operation data set based on the user's operation on the target client corresponding to the air conditioner, and / or, based on the target operation stage of the air conditioner corresponding to the operation data set, extract features from the operation data set according to the target feature extraction method, and obtain the operation feature set corresponding to the air conditioner, wherein the operation feature set includes the operation features of the air conditioner at multiple sampling moments, and perform fault diagnosis on the air conditioner based on the operation feature set. Thus, the target feature extraction method can be determined to extract features from the operation data set in consideration of the user's operation on the interactive page and / or the target operation stage of the air conditioner corresponding to the operation data set, thereby improving the accuracy of the target feature extraction method, facilitating the collection of features that are relatively small in number and strongly related to the fault diagnosis of the air conditioner, and automatically performing fault diagnosis on the air conditioner using the operation feature set, thereby improving the accuracy and efficiency of the fault diagnosis of the air conditioner and saving computing resources.
[0271] In order to implement the above embodiments, the present disclosure further proposes a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, implement the steps of the air conditioner fault diagnosis method and model training method provided by the present disclosure.
[0272] Alternatively, the computer readable storage medium may be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, or the like.
[0273] In order to implement the above embodiments, the present disclosure also provides a computer program product, including a computer program, characterized in that when the computer program is executed by a processor of an electronic device, the fault diagnosis method and model training method of the air conditioner as described above are implemented.
[0274] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The description and examples are to be considered exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.
[0275] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.
Claims
1. A method for diagnosing a fault of an air conditioner, characterized in that: include: Acquire an operation data set corresponding to the air conditioner, wherein the operation data set includes operation data of the air conditioner at multiple sampling times; Determining a target feature extraction method for the operation data set based on an operation of a user on a target client corresponding to the air conditioner and / or based on a target operation stage of the air conditioner corresponding to the operation data set; According to the target feature extraction method, feature extraction is performed on the operation data set to obtain an operation feature set corresponding to the air conditioner, wherein the operation feature set includes operation features of the air conditioner at multiple sampling times; Based on the operating feature set, fault diagnosis is performed on the air conditioner.
2. The method according to claim 1, characterized in that The method of extracting the target features of the operating data set based on the operation of the user on the target client corresponding to the air conditioner and / or based on the target operating stage of the air conditioner corresponding to the operating data set comprises: Controlling display of a plurality of first candidate selection items on the interactive page of the target client, wherein the first candidate selection items are used to represent candidate feature extraction methods; Based on the selection operation of the user on the interactive page for the first target option, the candidate feature extraction method corresponding to the first target option is used as the target feature extraction method.
3. The method according to claim 1, characterized in that The step of obtaining the operating data set corresponding to the air conditioner includes: Controlling display of a plurality of second candidate selection items and a plurality of third candidate selection items on the interactive page of the target client, wherein the second candidate selection items are used to represent candidate sampling time periods, and the third candidate selection items are used to represent candidate categories of the operating data of the air conditioner; Based on the selection operation of the user on the second target option on the interactive page, taking the candidate sampling time period corresponding to the second target option as the target sampling time period; Based on the selection operation of the user on the third target option on the interactive page, taking the candidate category corresponding to the third target option as the target category; The operation data of the target category of the air conditioner in the target sampling time period is added to the operation data set.
4. The method according to claim 1, characterized in that Before obtaining the operating data set corresponding to the air conditioner, the method further includes: An activation operation of the user for a fault diagnosis function on the interactive page of the target client is detected.
5. The method according to claim 1, characterized in that The method of extracting the target features of the operating data set based on the operation of the user on the target client corresponding to the air conditioner and / or based on the target operating stage of the air conditioner corresponding to the operating data set comprises: Based on the target operation stage, determining the target feature category to be extracted from the operation data set as the target feature extraction method; The step of extracting features from the operating data set according to the target feature extraction method to obtain an operating feature set corresponding to the air conditioner includes: Performing feature extraction on the operation data set to obtain operation features of the target feature category; Adding the operation features of the target feature category to the operation feature set.
6. The method according to claim 5, characterized in that If there are multiple target operation stages, determining the target feature category to be extracted from the operation data set based on the target operation stage includes: Splitting the operation data set to obtain a plurality of operation sub-data sets, wherein the operation sub-data sets correspond one-to-one to the target operation stage; Determining a sub-feature category to be extracted from the operation sub-dataset based on a target operation stage corresponding to the operation sub-dataset; The extracting features of the operation data set to obtain the operation features of the target feature category includes: Feature extraction is performed on the operation sub-dataset to obtain the operation features of the sub-feature category.
7. The method according to claim 5, characterized in that The step of determining the target feature category to be extracted from the operating data set based on the target operating stage includes: Acquire a mapping relationship between candidate operating stages of the air conditioner and candidate feature categories; A candidate feature category having a mapping relationship with the target operating stage is obtained as the target feature category.
8. The method according to claim 5, characterized in that The target feature category includes at least one of a dynamic feature category, a steady-state feature category, and a cumulative feature category.
9. The method according to claim 8, characterized in that The step of determining the target feature category to be extracted from the operating data set based on the target operating stage includes: If the target operation stage includes a dynamic stage, determining that the target feature category includes the dynamic feature category and the cumulative feature category; If the target operation stage includes a steady-state stage, it is determined that the target feature category includes the steady-state feature category and the cumulative feature category.
10. The method according to claim 8, characterized in that The dynamic feature category includes a change speed and / or a change acceleration of the operating data of the air conditioner.
11. The method according to claim 1, characterized in that: Obtaining the target operation stage includes: If the operation data set includes operation data at a first moment, wherein the plurality of sampling moments include the first moment, and the accumulated operation time of the air conditioner at the first moment is less than or equal to a set time, it is determined that the target operation stage includes a dynamic stage; If the operating data set includes operating data at a second moment, wherein the plurality of sampling moments include the second moment, and the accumulated operating time of the air conditioner at the second moment is greater than the set time, it is determined that the target operating stage includes a steady-state stage.
12. The method according to claim 1, characterized in that Obtaining the target operation stage includes: If the operation data set includes operation data at a third moment, wherein the plurality of sampling moments include the third moment, and the absolute value of the difference between the indoor ambient temperature and the set temperature at the third moment is greater than a set threshold, it is determined that the target operation stage includes a dynamic stage; If the operating data set includes operating data at a fourth moment, wherein the plurality of sampling moments include the fourth moment, and the absolute value of the difference at the fourth moment is less than or equal to the set threshold, it is determined that the target operating stage includes a steady-state stage.
13. The method according to any one of claims 1 to 12, characterized in that The method of performing fault diagnosis on the air conditioner based on the operation feature set includes: The operating feature set is input into a target fault diagnosis model, and the target fault diagnosis model outputs a fault diagnosis result of the air conditioner.
14. The method according to claim 13, characterized in that The method further comprises: Controlling to display a plurality of fourth candidate selection items on the interactive page of the target client, wherein the fourth candidate selection items are used to represent the candidate fault diagnosis model; Based on the selection operation of the user on the interactive page for the fourth target selection item, the candidate fault diagnosis model corresponding to the fourth target selection item is used as the target fault diagnosis model.
15. The method according to any one of claims 1 to 12, characterized in that The method further comprises: Controlling to display the fault diagnosis data of the air conditioner on the interactive page of the target client, wherein the fault diagnosis data includes the target feature extraction method and / or the fault diagnosis result of the air conditioner.
16. A model training method, characterized in that: include: Acquire a sample operation data set corresponding to the sample air conditioner, wherein the sample operation data set includes sample operation data of the sample air conditioner at multiple sampling moments; Determine a sample feature extraction method of the sample operation data set based on an operation of a sample user on a sample client corresponding to the sample air conditioner and / or based on a sample operation stage of the sample air conditioner corresponding to the sample operation data set; According to the sample feature extraction method, feature extraction is performed on the sample operation data set to obtain a sample operation feature set corresponding to the sample air conditioner, wherein the sample operation feature set includes sample operation features of the sample air conditioner at multiple sampling times; The fault diagnosis model is trained based on the sample running feature set.
17. The method according to claim 16, characterized in that The method of extracting sample features of the sample operation data set based on the operation of the sample user on the sample client corresponding to the sample air conditioner and / or based on the sample operation stage of the sample air conditioner corresponding to the sample operation data set comprises: Controlling display of a plurality of first candidate selection items on the interactive page of the sample client, wherein the first candidate selection items are used to represent candidate feature extraction methods; Based on the selection operation of the sample user on the interactive page for the first target option, the candidate feature extraction method corresponding to the first target option is used as the sample feature extraction method.
18. The method according to claim 16, characterized in that The method of extracting sample features of the sample operation data set based on the operation of the sample user on the sample client corresponding to the sample air conditioner and / or based on the sample operation stage of the sample air conditioner corresponding to the sample operation data set comprises: Based on the sample running stage, determining a sample feature category to be extracted from the sample running data set as the sample feature extraction method; The method of extracting features from the sample operation data set according to the sample feature extraction method to obtain a sample operation feature set corresponding to the sample air conditioner includes: Performing feature extraction on the sample operation data set to obtain sample operation features of the sample feature category; Adding the sample running features of the sample feature category to the sample running feature set.
19. The method according to claim 16, characterized in that The sample operation data set also includes sample fault diagnosis results of the sample air conditioner. The training of the fault diagnosis model based on the sample operation feature set includes: Inputting the sample operation feature set into the fault diagnosis model, and having the fault diagnosis model output a predicted fault diagnosis result of the sample air conditioner; The fault diagnosis model is trained based on the predicted fault diagnosis results of the sample air conditioners and the sample fault diagnosis results of the sample air conditioners.
20. A fault diagnosis device for an air conditioner, characterized in that: include: An acquisition module is configured to acquire an operation data set corresponding to the air conditioner, wherein the operation data set includes operation data of the air conditioner at multiple sampling moments; a determination module configured to determine a target feature extraction method of the operation data set based on an operation of a user on a target client corresponding to the air conditioner and / or based on a target operation stage of the air conditioner corresponding to the operation data set; a feature extraction module configured to perform feature extraction on the operation data set according to the target feature extraction method to obtain an operation feature set corresponding to the air conditioner, wherein the operation feature set includes operation features of the air conditioner at multiple sampling moments; The fault diagnosis module is configured to perform fault diagnosis on the air conditioner based on the operation feature set.
21. A model training device, characterized in that: include: An acquisition module is configured to acquire a sample operation data set corresponding to the sample air conditioner, wherein the sample operation data set includes sample operation data of the sample air conditioner at multiple sampling moments; a determination module configured to determine a sample feature extraction method of the sample operation data set based on an operation of a sample user on a sample client corresponding to the sample air conditioner and / or based on a sample operation stage of the sample air conditioner corresponding to the sample operation data set; a feature extraction module configured to perform feature extraction on the sample operation data set according to the sample feature extraction method to obtain a sample operation feature set corresponding to the sample air conditioner, wherein the sample operation feature set includes sample operation features of the sample air conditioner at multiple sampling moments; The training module is configured to execute training on the fault diagnosis model based on the sample running feature set.
22. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; Wherein, the processor is configured to: Implement the steps of the method according to any one of claims 1 to 19.
23. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the program instructions are executed by a processor, the steps of the method described in any one of claims 1 to 19 are implemented.
