Fault identification method and device of heating ventilation air conditioner, computer equipment and storage medium

Abstract

The invention relates to a fault identification method and device for a heating ventilation air conditioner, computer equipment and a storage medium. The method comprises the steps that operation parameters of a target heating ventilation air conditioner are obtained; a preset fault recognition model is used for conducting fault recognition on the operation parameters, and fault information of the target heating ventilation air conditioner in the operation process is determined; wherein the preset fault identification model is obtained by training a training result of original sample operation parameters by using sample migration operation parameters in different migration scenes; the sample migration operation parameters in different migration scenes are operation parameters corresponding to different operation conditions and different air conditioner types predicted based on the sample original operation parameters. By adopting the method, faults of a complex system in a changeable operation environment can be accurately identified.

Description

Technical Field

[0001] The present application relates to the technical field of fault identification, and in particular to a method, device, computer equipment and storage medium for identifying faults of a heating, ventilation and air conditioning system. Background Art

[0002] In modern building management systems, fault diagnosis of Heating, Ventilation and Air Conditioning (HVAC) systems has become a key factor in improving building energy efficiency and ensuring operational safety.

[0003] In the related technology, the air-conditioning operation data is mainly analyzed based on deep learning to determine whether there is a fault during the operation of the air-conditioning.

[0004] However, when faced with complex systems and changing operating environments, related technical methods often have difficulty in accurately and quickly identifying faults due to their lack of flexibility and adaptability. Summary of the invention

[0005] Based on this, it is necessary to provide a HVAC fault identification method, device, computer equipment and storage medium to address the above technical problems, which can accurately identify faults in complex systems and changing operating environments.

[0006] In a first aspect, the present application provides a method for identifying a fault of a heating, ventilation and air conditioning system, comprising:

[0007] Obtaining the operating parameters of the target HVAC;

[0008] Use the preset fault identification model to identify the operating parameters and determine the fault information of the target HVAC during operation;

[0009] Among them, the preset fault identification model is obtained by training the training results of the sample original operating parameters using the sample migration operating parameters under different migration scenarios; the sample migration operating parameters under different migration scenarios are based on the sample original operating parameters to predict the operating parameters corresponding to different operating conditions and different air-conditioning types.

[0010] In one embodiment, the fault identification model includes a general fault model and a migration fault identification model; using the preset fault identification model to perform fault identification on the operating parameters to determine the fault information of the target HVAC during operation, including:

[0011] Performing universal fault identification on the operating parameters through the universal fault model to obtain a first fault identification result;

[0012] Identify the migration fault of the operating parameter by using the migration fault identification model to obtain a second fault identification result;

[0013] Based on the first fault identification result and the second fault identification result, fault information of the target HVAC during operation is determined.

[0014] In one embodiment, the process of constructing the fault identification model includes:

[0015] Obtaining sample original operating parameters of different types of HVAC during operation; and obtaining sample migration operating parameters under different migration scenarios; the sample original operating parameters and the sample migration operating parameters both include normal operating parameters and fault operating parameters;

[0016] The initial fault recognition model is trained by using the original operating parameters of the samples to obtain an intermediate fault recognition model;

[0017] The intermediate fault recognition model is trained by using sample migration operation parameters under different migration scenarios to obtain a fault recognition model.

[0018] In one embodiment, the intermediate fault identification model is trained by using sample migration operation parameters in different migration scenarios to obtain a fault identification model, including:

[0019] The intermediate fault recognition model is processed by different transfer learning methods to obtain multiple initial transfer fault recognition models;

[0020] The initial migration fault identification models are trained respectively by using sample migration operation parameters under different migration scenarios to obtain multiple migration fault identification models;

[0021] Based on each migration fault identification model, a fault identification model is determined.

[0022] In one embodiment, determining a fault identification model based on each migration fault identification model includes:

[0023] Evaluate multiple migration fault identification models respectively to obtain evaluation results of each migration fault identification model;

[0024] The migration fault identification model with the best evaluation result is selected from each evaluation result as the fault identification model.

[0025] In one embodiment, multiple migration fault identification models are evaluated respectively to obtain evaluation results of each migration fault identification model, including:

[0026] For any migration fault identification model, performance evaluation parameters are obtained from sample migration operation parameters under different migration scenarios;

[0027] Performing fault identification on the performance evaluation parameter by using the migration fault identification model to obtain a first fault test result; and performing fault identification on the performance evaluation parameter by using the intermediate fault identification model to obtain a second fault test result;

[0028] An evaluation result corresponding to the migration fault identification model is determined according to the first fault test result and the second fault test result.

[0029] In one embodiment, determining an evaluation result corresponding to the migration fault identification model according to the first fault test result and the second fault test result includes:

[0030] Obtain a fault test error between a first fault test result and a second fault test result;

[0031] The fault test error is evaluated based on the preset evaluation rules to obtain the evaluation result corresponding to the migration fault identification model.

[0032] In a second aspect, the present application also provides a fault identification device for a heating, ventilation and air conditioning system, comprising:

[0033] An acquisition module, used to acquire the operating parameters of the target HVAC;

[0034] An identification module, used to identify faults of operating parameters using a preset fault identification model, and determine fault information of the target HVAC during operation;

[0035] Among them, the preset fault identification model is obtained by training the training results of the sample original operating parameters using the sample migration operating parameters under different migration scenarios; the sample migration operating parameters under different migration scenarios are based on the sample original operating parameters to predict the operating parameters corresponding to different operating conditions and different air-conditioning types.

[0036] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the content of any one embodiment of the HVAC fault identification method in the above-mentioned first aspect is implemented.

[0037] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the content of any one embodiment of the HVAC fault identification method in the first aspect above.

[0038] In a fifth aspect, the present application further provides a computer program product, including a computer program, which, when executed by a processor, implements the content of any one embodiment of the HVAC fault identification method in the above-mentioned first aspect.

[0039] The above-mentioned HVAC fault identification method, device, computer equipment and storage medium obtain the operating parameters of the target HVAC; use the preset fault identification model to identify the operating parameters and determine the fault information of the target HVAC during operation; wherein the preset fault identification model is obtained by training the training results of the sample original operating parameters using the sample migration operating parameters under different migration scenarios; the sample migration operating parameters under different migration scenarios are based on the sample original operating parameters to predict the operating parameters corresponding to different operating conditions and different air conditioning types. The fault identification model in this method is obtained by training the training results of the sample original operating parameters using the sample migration operating parameters under different migration scenarios, that is, the fault identification model can not only learn the characteristics of the original operating parameters, but also learn the characteristics of the sample migration operating parameters under different migration scenarios, which enhances the adaptability of the fault identification model to the unseen HVAC system types and operating conditions, and expands the application scope of the fault identification model. Then, no matter what type of target HVAC, the fault identification model can accurately analyze the operating parameters of the target HVAC and accurately determine the fault information of the target HVAC during operation. The fault identification model can be used to flexibly analyze the operating parameters of different types and working conditions, which can better meet the needs of practical applications. At the same time, the sample migration operating parameters under different migration scenarios are based on the original sample operating parameters to predict the operating parameters corresponding to different operating conditions and different air-conditioning types. In other words, based on the original small amount of data, the sample migration operating parameters under different migration scenarios can be obtained through prediction, which can expand the training sample data of the model and reduce the dependence on a large amount of labeled data, so that the fault identification model can still maintain efficient fault diagnosis performance when data acquisition is difficult or costly. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 A diagram of an application environment of a fault identification method for HVAC in one embodiment;

[0042] Figure 2 A schematic diagram of a flow chart of a method for identifying a fault of a heating, ventilation and air conditioning system in one embodiment;

[0043] Figure 3 A schematic diagram of a flow chart of a method for identifying a fault of a heating, ventilation and air conditioning system in one embodiment;

[0044] Figure 4 A schematic diagram of a flow chart of a method for identifying a fault of a heating, ventilation and air conditioning system in one embodiment;

[0045] Figure 5 A schematic diagram of a flow chart of a method for identifying a fault of a heating, ventilation and air conditioning system in one embodiment;

[0046] Figure 6 A schematic diagram of a flow chart of a method for identifying a fault of a heating, ventilation and air conditioning system in one embodiment;

[0047] Figure 7 A schematic diagram of a flow chart of a method for identifying a fault of a heating, ventilation and air conditioning system in one embodiment;

[0048] Figure 8 A schematic diagram of test results of multiple migration fault identification models in one embodiment;

[0049] Fig. 9 A schematic diagram of test results of multiple migration fault identification models in one embodiment;

[0050] Fig.10 A fault identification accuracy variation curve diagram of multiple migration fault identification models in one embodiment;

[0051] Fig.11 A fault identification accuracy variation curve diagram of multiple migration fault identification models in one embodiment;

[0052] Fig.12 A network structure diagram of a network-based deep transfer learning strategy in one embodiment;

[0053] Fig.13 A schematic diagram of a flow chart of a method for identifying a fault of a heating, ventilation and air conditioning system in one embodiment;

[0054] Fig.14 A schematic diagram of a training process of a fault identification model in one embodiment;

[0055] Fig.15 A schematic diagram of a training process of a fault identification model for HVAC in one embodiment;

[0056] Fig.16 A schematic diagram of a flow chart of a method for identifying a fault of a heating, ventilation and air conditioning system in one embodiment;

[0057] Fig.17 is a structural block diagram of a fault identification device for a heating, ventilation and air conditioning system in one embodiment;

[0058] Fig.18 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0060] Before introducing the technical solution of the present application in detail, a brief description of the background technology of the present application is first given.

[0061] In modern building management systems, the energy efficiency and fault diagnosis of HVAC systems have become key factors in improving building energy efficiency and ensuring operational safety. Traditional fault diagnosis methods mostly rely on expert systems and rule-based analysis. Although these methods are effective under specific conditions, they often lack flexibility and adaptability and are difficult to accurately and quickly identify faults in complex systems and changing operating environments. In recent years, with the development of data-driven methods and machine learning technologies, data-based fault diagnosis methods have received widespread attention. In particular, deep learning models have been shown to excel in fault diagnosis tasks due to their excellent feature extraction and pattern recognition capabilities.

[0062] However, existing fault diagnosis methods based on deep learning still face some challenges and limitations. First, the training of deep learning models usually requires a large amount of labeled data. In practical applications, it is difficult to obtain sufficient fault sample data, especially for newly installed HVAC systems or those with rare fault types, which limits the training and optimization of the model. Second, since the design and operating parameters of different HVAC systems may vary greatly, and the data distribution of the same system under different working conditions may also be different, existing models are often difficult to adapt to new system types or new operating conditions, resulting in a decrease in fault diagnosis accuracy, and the generalization ability and scope of application of the model are limited.

[0063] In response to the above problems, the present application provides a HVAC fault identification method, device, computer equipment and storage medium, which can not only use limited labeled data to efficiently train an accurate fault identification model, but also improve the model's adaptability and accuracy to new systems and new working conditions by migrating the knowledge learned in different systems and working conditions, thereby effectively expanding the scope of application of the model and improving the accuracy of fault diagnosis. This is to make up for the shortcomings of the air conditioning system fault diagnosis method based on deep learning in the prior art in terms of low data availability, poor model generalization ability and limited scope of application. Of course, the technical solution provided in the embodiments of the present application is not limited to solving only the above problems, but also has other technical effects, which can be specifically referred to in the following embodiment description. The technical solution of the present application is described in detail below.

[0064] The HVAC fault identification method provided in the embodiment of the present application can be applied to Figure 1 The application environment shown. The application environment includes a computer device 101 and a target HVAC 102. The computer device 101 and the target HVAC 102 can communicate remotely, or the computer device 101 can be integrated into the target HVAC control 102. The computer device 101 is used to obtain the operating parameters of the target HVAC 102, and use a preset fault identification model to perform fault identification on the operating parameters to determine the fault information of the target HVAC 102 during operation.

[0065] In an exemplary embodiment, Figure 2 As shown, a fault identification method for HVAC is provided, and the method is applied to Figure 1 The computer device in the example is used to illustrate, including the following steps 101 to 102. Among them:

[0066] S101, obtaining operating parameters of a target HVAC system.

[0067] The target HVAC may be any type of HVAC, including but not limited to screw chillers, centrifugal chillers, etc. The operating parameters may be parameters such as air volume, air temperature, air humidity, and air pressure of the target HVAC.

[0068] In the embodiment of the present application, during the operation of the target HVAC, the computer device may send a parameter acquisition instruction to the target HVAC, and when the target HVAC receives the parameter acquisition instruction sent by the computer device, the operating parameters required in the parameter acquisition instruction may be sent to the computer device. For example, the operating parameters required in the parameter acquisition instruction may be the operating parameters of the target HVAC at the current moment, or may also be the operating parameters of the target HVAC at a historical moment.

[0069] Alternatively, during the operation of the target HVAC, the target HVAC may also send current operating parameters to the computer device at preset time intervals.

[0070] Optionally, when the target HVAC is not in operation, the historical operating parameters of the target HVAC are stored in a parameter memory. The computer device can use the identification information of the target HVAC to search the parameter memory for operating parameters that match the identification information. The embodiment of the present application does not limit the method for obtaining the operating parameters of the target HVAC.

[0071] S102, using a preset fault identification model to perform fault identification on the operating parameters to determine fault information of the target HVAC during operation;

[0072] Among them, the preset fault identification model is obtained by training the training results of the sample original operating parameters using the sample migration operating parameters under different migration scenarios; the sample migration operating parameters under different migration scenarios are based on the sample original operating parameters to predict the operating parameters corresponding to different operating conditions and different air-conditioning types.

[0073] The fault information may include fault time, fault type, fault severity and other information.

[0074] During the fault identification process, the computer equipment can input the operating parameters of the target HVAC into a preset fault identification model. The preset fault identification model can be used to perform fault analysis on the operating parameters of the target HVAC to determine whether the target HVAC has a fault during operation and related information about the fault.

[0075] In the training process of the fault identification model, the initial fault identification model is first preliminarily trained through the training results of the original operating parameters of the samples to obtain the preliminary training results. Furthermore, different operating conditions and different air conditioner types are added on the basis of the original operating parameters of the samples to simulate the operating parameters corresponding to different operating conditions and different air conditioner types, that is, the sample migration operating parameters under different migration scenarios. The preliminary training results are further trained using the sample migration operating parameters under different migration scenarios to obtain the fault identification model.

[0076] In the above-mentioned HVAC fault identification method, the operating parameters of the target HVAC are obtained; the operating parameters are fault identified using a preset fault identification model to determine the fault information of the target HVAC during operation; wherein the preset fault identification model is obtained by training the training results of the sample original operating parameters using the sample migration operating parameters under different migration scenarios; the sample migration operating parameters under different migration scenarios are based on the sample original operating parameters to predict the operating parameters corresponding to different operating conditions and different air conditioning types. The fault identification model in this method is obtained by training the training results of the sample original operating parameters using the sample migration operating parameters under different migration scenarios, that is, the fault identification model can not only learn the characteristics of the original operating parameters, but also learn the characteristics of the sample migration operating parameters under different migration scenarios, which enhances the adaptability of the fault identification model to unseen HVAC system types and operating conditions, and expands the application scope of the fault identification model. Therefore, no matter what type of target HVAC, the fault identification model can accurately analyze the operating parameters of the target HVAC and accurately determine the fault information of the target HVAC during operation. The fault identification model can be used to flexibly analyze the operating parameters of different types and working conditions, which can better meet the needs of practical applications. At the same time, the sample migration operating parameters under different migration scenarios are based on the original sample operating parameters to predict the operating parameters corresponding to different operating conditions and different air-conditioning types. In other words, based on the original small amount of data, the sample migration operating parameters under different migration scenarios can be obtained through prediction, which can expand the training sample data of the model and reduce the dependence on a large amount of labeled data, so that the fault identification model can still maintain efficient fault diagnosis performance when data acquisition is difficult or costly.

[0077] Assume that the fault identification model includes a general fault model and a migration fault identification model; in one embodiment, Figure 3 As shown, the specific content of the above-mentioned method of using the preset fault identification model to identify the operating parameters to determine the fault information of the target HVAC during operation is introduced, and the specific content includes the following steps:

[0078] S201, performing general fault identification on operating parameters using a general fault model to obtain a first fault identification result.

[0079] Among them, common faults refer to faults that may exist in different HVAC systems and HVAC under different working conditions.

[0080] In the embodiment of the present application, the fault identification model is composed of two models, a general fault model and a migration fault identification model, and the general fault model is set at the front end of the migration fault identification model. The computer device inputs the operating parameters of the target HVAC into the fault identification model, and the operating parameters first enter the general fault model. The general fault model extracts the characteristic information of the operating parameters, and analyzes whether the operating parameters have faults and corresponding fault-related information based on the characteristic information of the operating parameters to obtain a first fault identification result.

[0081] S202: Identify the migration fault of the operating parameter by using a migration fault identification model to obtain a second fault identification result.

[0082] Among them, migration faults refer to some fault types that are unique to HVAC in different HVAC systems and under different working conditions.

[0083] In an embodiment of the present application, after the operating parameters are identified in the general fault identification model, the operating parameters enter the migration fault identification model. The migration fault identification model can identify the migration fault for the operating parameters, determine whether there is a migration fault and corresponding fault information in the operating parameters, and obtain a second fault identification result.

[0084] S203: Determine fault information of the target HVAC during operation based on the first fault identification result and the second fault identification result.

[0085] In the embodiment of the present application, the first fault identification result is the result obtained by performing general fault identification, and the second fault identification result is the result obtained by performing migration fault identification. Since general faults and migration faults are different types of faults, after obtaining the first fault identification result and the second fault identification result, the computer device can merge the first fault identification result and the second fault identification result to obtain the fault information of the target HVAC during operation.

[0086] In the above-mentioned HVAC fault identification method, a general fault identification is performed on the operating parameters through a general fault model to obtain a first fault identification result; a migration fault of the operating parameters is identified through a migration fault identification model to obtain a second fault identification result; based on the first fault identification result and the second fault identification result, the fault information of the target HVAC during operation is determined. In the fault identification process, this method uses a general fault model for general fault identification and a migration fault identification model for migration fault identification, which can perform more comprehensive fault identification from different angles, so that the fault information of the target HVAC during operation will be more accurate.

[0087] The above embodiments are all introductions to the application process of fault identification. Next, the construction process of the fault identification model is described in detail. Figure 4 As shown, the construction process of the above fault identification model includes the following steps:

[0088] S301, obtaining sample original operating parameters of different types of HVAC during operation; and obtaining sample migration operating parameters under different migration scenarios; the sample original operating parameters and the sample migration operating parameters both include normal operating parameters and fault operating parameters.

[0089] In an embodiment of the present application, the computer device can obtain the operating parameters of different types of HVAC in the historical operation process from the parameter storage, and use the operating parameters in the historical operation process as the sample original operating parameters. Further, the computer device can predict the sample migration operating parameters under different migration scenarios based on the sample original operating parameters. Among them, different migration scenarios refer to the operating parameters corresponding to different operating conditions and different HVAC types.

[0090] S302, training an initial fault identification model using original operating parameters of the sample to obtain an intermediate fault identification model.

[0091] In the embodiment of the present application, during the training process, the computer device can input the original operating parameters of the sample into the initial fault identification model to obtain a preliminary identification result, and adjust the parameters of the initial fault identification model based on the error between the preliminary identification result and the gold standard corresponding to the original operating parameters of the sample, until the error between the result output by the initial fault identification model and the gold standard corresponding to the original operating parameters of the sample is small, then the training process is completed and the intermediate fault identification model is obtained. It can be understood that the intermediate fault identification model has learned common faults of many different types of HVAC under different operating conditions.

[0092] S303, training the intermediate fault identification model by using sample migration operation parameters in different migration scenarios to obtain a fault identification model.

[0093] In the embodiment of the present application, after obtaining the intermediate fault identification model, the computer device can input the sample migration operation parameters under different migration scenarios into the intermediate fault identification model, use the intermediate fault identification model to analyze the sample migration operation parameters under different migration scenarios, obtain the identification result, and adjust the parameters of the intermediate fault identification model based on the identification result. Until the identification result output by the intermediate fault identification model meets the preset conditions, the fault identification model is obtained.

[0094] In the above-mentioned HVAC fault identification method, sample original operating parameters of different types of HVAC during operation are obtained; and sample migration operating parameters under different migration scenarios are obtained; the sample original operating parameters and the sample migration operating parameters both include normal operating parameters and fault operating parameters; the initial fault identification model is trained by the sample original operating parameters to obtain an intermediate fault identification model; the intermediate fault identification model is trained by the sample migration operating parameters under different migration scenarios to obtain a fault identification model. This method uses the sample original operating parameters to train the initial fault identification model. During this training process, the initial fault identification model can learn common faults. Based on this, the intermediate fault identification model obtained by the above training is further trained using the sample migration operating parameters under different migration scenarios, and can further learn migration faults, thereby enhancing the adaptability of the fault identification model to unseen HVAC system types and operating conditions, and expanding the application scope of the fault identification model.

[0095] Next, the specific contents of obtaining the fault identification model by training the intermediate fault identification model through the sample migration operation parameters in different migration scenarios are described through an embodiment. Figure 5 As shown, the specific content includes:

[0096] S401, processing the intermediate fault identification model by different transfer learning methods to obtain multiple initial transfer fault identification models.

[0097] In the embodiment of the present application, after the intermediate fault identification model is obtained, the computer device can adjust the hyperparameters and structure of the intermediate fault identification model, and analyze the adjusted intermediate fault identification model using different transfer learning methods to obtain multiple initial transfer fault identification models. For example, the initial transfer fault identification model can include network-based deep transfer learning (FT), mapping-based deep transfer learning (DaNN), and adversarial deep transfer learning (DANN).

[0098] S402 , respectively training each initial migration fault identification model by using sample migration operation parameters in different migration scenarios to obtain a plurality of migration fault identification models.

[0099] In an embodiment of the present application, after obtaining multiple initial migration fault identification models, the computer device can use sample migration operation parameters under different migration scenarios to train each initial migration fault identification model separately to obtain a migration fault identification model corresponding to each initial migration fault identification model, that is, multiple migration fault identification models.

[0100] S403: Determine a fault identification model based on each migration fault identification model.

[0101] In an embodiment of the present application, after obtaining multiple migration fault identification models, the computer device can evaluate each migration fault identification model using preset evaluation rules, or test and evaluate each migration fault identification model using test data, and use the migration fault identification model with the highest evaluation result as the fault identification model.

[0102] In the above-mentioned HVAC fault identification method, the intermediate fault identification model is processed by different transfer learning methods to obtain multiple initial migration fault identification models; each initial migration fault identification model is trained by sample migration operation parameters under different migration scenarios to obtain multiple migration fault identification models; based on each migration fault identification model, a fault identification model is determined. This method processes the intermediate fault identification model by transfer learning, and can obtain multiple initial migration fault identification models, and then trains multiple initial migration fault identification models simultaneously by sample migration operation parameters. Based on the training results, the fault identification model with the best fault identification effect can be accurately found from multiple migration fault identification models to improve the identification accuracy of the fault identification model.

[0103] In one embodiment, Figure 6 As shown, the specific content of determining the fault identification model based on each migration fault identification model is described, and the specific content includes:

[0104] S501 , respectively evaluating a plurality of migration fault identification models to obtain evaluation results of the respective migration fault identification models.

[0105] In the embodiment of the present application, after obtaining multiple migration fault identification models, the computer device can evaluate each migration fault identification model using a preset evaluation rule to obtain an evaluation result of each migration fault identification model. Alternatively, the computer device can also test each migration fault identification model using test data, compare the obtained multiple test results with the preset test results, and determine the evaluation result of each migration fault identification model based on the comparison result.

[0106] S502 , selecting a migration fault identification model with the best evaluation result from each evaluation result as a fault identification model.

[0107] In an embodiment of the present application, the computer device may obtain a migration fault identification model corresponding to the best evaluation result selected from multiple evaluation results, and use the migration fault identification model corresponding to the best evaluation result as the fault identification model.

[0108] In the above-mentioned fault identification method for HVAC, multiple migration fault identification models are evaluated respectively to obtain evaluation results of each migration fault identification model; and the migration fault identification model with the best evaluation result is selected from each evaluation result as the fault identification model. By evaluating each migration fault identification model, the method can quantify the fault identification performance of each migration fault identification model in a quantitative manner, so that by comparing multiple quantification results, a fault identification model can be quickly and accurately screened out from multiple migration fault identification models.

[0109] In one embodiment, Figure 7 As shown, the specific contents of the evaluation results of each migration fault identification model obtained by respectively evaluating the multiple migration fault identification models include:

[0110] S601 , for any migration fault identification model, obtaining performance evaluation parameters from sample migration operation parameters in different migration scenarios.

[0111] In an embodiment of the present application, the computer device can divide the sample migration operation parameters in different migration scenarios into training parameters and other parameters except the training parameters according to a preset ratio, and use the other parameters except the training parameters as performance evaluation parameters. For example, the preset ratio can be 7:3, that is, the training parameters account for 70% of the sample migration operation parameters, and the performance evaluation parameters account for 30% of the sample migration operation parameters.

[0112] S602, performing fault identification on the performance evaluation parameter by using a migration fault identification model to obtain a first fault test result; and performing fault identification on the performance evaluation parameter by using an intermediate fault identification model to obtain a second fault test result.

[0113] In an embodiment of the present application, after obtaining the performance evaluation parameters, the computer device can use the performance evaluation parameters as test data to test the migration fault identification model and the intermediate fault identification model respectively, and determine the evaluation result corresponding to the migration fault identification model based on the test results. Specifically, the computer device can input the performance evaluation parameters into the migration fault identification model, use the migration fault identification model to perform fault identification on the performance evaluation parameters, and obtain a first fault test result. The computer device can also input the performance evaluation parameters into the intermediate fault identification model, use the intermediate fault identification model to perform fault identification on the performance evaluation parameters, and obtain a second fault test result.

[0114] S603: Determine an evaluation result corresponding to the migration fault identification model according to the first fault test result and the second fault test result.

[0115] In an embodiment of the present application, after obtaining the first fault test result and the second fault test result, the computer device can calculate the fault test difference between the first fault test result and the second fault test result, and based on the correspondence between the fault test difference and the evaluation value, map the fault test difference to obtain the evaluation result corresponding to the migration fault identification model.

[0116] Assuming that centrifugal chillers are used as training data and screw chillers are used as test data, after testing the test data, the fault identification model obtained by the network-based deep transfer learning strategy has a high fault diagnosis accuracy under different migration tasks. Especially in the cross-system migration task, the fault diagnosis accuracy of the network-based deep transfer learning strategy is significantly higher than that of the baseline model, which fully demonstrates the effectiveness and superiority of the network-based deep transfer learning strategy in solving cross-condition and cross-system HVAC fault diagnosis.

[0117] Figure 8 This is a schematic diagram of the test results of multiple migration fault identification models. The figure includes six groups of test results. Each group of test results includes the test results of different migration fault identification models and an intermediate fault identification model. The test results of different groups indicate that the test result of the migration fault identification model corresponding to the network-based deep transfer learning strategy is the best. Fig. 9 This is a schematic diagram of the test results of multiple migration fault identification models, which includes two groups of test results. From the two groups of test results, it can be seen that the test result of the migration fault identification model corresponding to the network-based deep transfer learning strategy is the best.

[0118] Fig.10 The fault recognition accuracy change curve of multiple migration fault recognition models is shown. It can be seen from the figure that with the increase of sample data volume, the test effect of the migration fault recognition model corresponding to the network-based deep transfer learning strategy is relatively stable and has a higher accuracy. Fig.11 It is also a curve diagram showing the change in fault recognition accuracy of multiple migration fault recognition models. It can be seen from the figure that as the amount of sample data increases, the test effect of the migration fault recognition model corresponding to the network-based deep transfer learning strategy increases rapidly, and the accuracy is higher than that of other migration fault recognition models. In other words, the migration fault recognition model corresponding to the network-based deep transfer learning strategy can effectively utilize the increased amount of data to improve the generalization ability of the model and the effect of transfer learning.

[0119] The core of the network-based deep transfer learning strategy is to use the sample migration operating parameters in different migration scenarios to fine-tune some network layers to adapt them to the data distribution of the sample migration operating parameters, thereby achieving more accurate fault identification. Fig.12A diagram showing the network structure of a network-based deep transfer learning strategy, which freezes the primary layers of the source domain pre-trained model, fine-tunes only the high-level layers, and trains using sample transfer operation parameters. This strategy retains the model's ability to recognize common features while improving its adaptability to specific sample transfer operation parameter features.

[0120] In the above-mentioned fault identification method for HVAC, for any migration fault identification model, performance evaluation parameters are obtained from sample migration operation parameters under different migration scenarios; fault identification is performed on the performance evaluation parameters through the migration fault identification model to obtain a first fault test result; and fault identification is performed on the performance evaluation parameters through the intermediate fault identification model to obtain a second fault test result; and the evaluation result corresponding to the migration fault identification model is determined according to the first fault test result and the second fault test result. For any migration fault identification model, the method uses the performance evaluation parameters in the sample migration operation parameters, and respectively uses the migration fault identification model and the intermediate fault identification model to evaluate the performance evaluation parameters, and takes the second fault test result of the intermediate fault identification model as a reference, so as to accurately evaluate the first fault test result of the migration fault identification model, thereby accurately obtaining the evaluation result corresponding to the migration fault identification model.

[0121] Next, the specific process of determining the evaluation result corresponding to the migration fault identification model according to the first fault identification result and the second fault identification result is described in detail through an embodiment. Fig.13 As shown, the process includes:

[0122] S701, obtaining a fault test error between a first fault test result and a second fault test result.

[0123] In the embodiment of the present application, the first fault test result and the second fault test result are both the results obtained by performing fault identification based on the performance test data. Since the second fault test result is obtained by performing fault identification on the performance evaluation parameter through the intermediate fault identification model, the second fault test result has the least fault data. The computer device can calculate the number of faults between the first fault test result and the second fault test result, and use the number of faults as the fault test error.

[0124] S702, evaluating the fault test error based on a preset evaluation rule to obtain an evaluation result corresponding to the migration fault identification model.

[0125] In the embodiment of the present application, the preset evaluation rule includes a rule between a fault test error and an evaluation value. After obtaining the fault test error between two fault test results, the computer device can use the preset evaluation rule to determine the evaluation value corresponding to the fault test error, and use the evaluation value corresponding to the fault test error as the evaluation result corresponding to the migration fault identification model.

[0126] In the above-mentioned HVAC fault identification method, a fault test error between the first fault test result and the second fault test result is obtained; the fault test error is evaluated based on a preset evaluation rule to obtain an evaluation result corresponding to the migration fault identification model. This method calculates the fault test error between the first fault test result and the second fault test result, and the fault test error represents the fault identification capability of the migration fault identification model. Then, the fault test error is evaluated using the evaluation rule, and the fault identification capability of the migration fault identification model can be accurately determined, thereby accurately determining the evaluation result corresponding to the migration fault identification model.

[0127] Fig.14 The figure is a schematic diagram of the training process of the fault identification model, which includes: obtaining the historical operation data of the HVAC, preprocessing the historical operation data, obtaining the sample original operation parameters, using the sample original operation parameters to train the general fault identification model, and after the training is completed, multiple initial migration fault identification models are obtained. Then, based on the sample original operation parameters, the sample migration operation parameters corresponding to different systems and different working conditions are obtained. The sample migration operation parameters are used to train multiple initial migration fault identification models respectively to obtain multiple migration fault identification models. Multiple migration fault identification models are evaluated, and based on the evaluation results, the best one is selected from the multiple migration fault identification models as the fault prediction model.

[0128] Fig.15 Figure 1 is a schematic diagram of the training process of the fault identification model. It can be seen from the figure that in the process of training the general fault identification model, the original operating parameters of the sample are used for training, and the network parameters of the initial general fault identification model are optimized to obtain the best general fault model. The hyperparameters and structure of the best general fault model are changed to obtain multiple initial migration fault identification models. The training data in the original operating parameters of the sample and the training data in the migration operating parameters of the sample are used to train multiple initial migration fault identification models respectively to obtain multiple migration fault identification models. The test data in the migration operating parameters of the sample are then used to test the multiple migration fault identification models. After that, the best general fault model is trained with the test data in the migration operating parameters of the sample to obtain a non-transfer learning baseline. The performance of multiple migration fault identification models is evaluated using the non-transfer learning baseline, and the model with the best effect among the multiple migration fault identification models is used as the fault identification model.

[0129] As a specific embodiment of the present application, the fault identification process of the HVAC is described in detail below. Fig.16 As shown, the HVAC fault identification method includes:

[0130] S801, obtaining operating parameters of a target HVAC system;

[0131] S802, performing general fault identification on the operating parameters using a general fault model to obtain a first fault identification result;

[0132] S803, identifying a migration fault of the operating parameter by using a migration fault identification model to obtain a second fault identification result;

[0133] S804: Determine fault information of the target HVAC during operation based on the first fault identification result and the second fault identification result.

[0134] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0135] Based on the same inventive concept, the embodiment of the present application also provides a HVAC fault identification device for implementing the HVAC fault identification method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the embodiments of one or more HVAC fault identification devices provided below can refer to the limitations of the HVAC fault identification method above, and will not be repeated here.

[0136] In an exemplary embodiment, Fig.17 As shown, a fault identification device for a heating, ventilation and air conditioning system is provided, comprising: an acquisition module 11 and an identification module 12, wherein:

[0137] An acquisition module 11 is used to acquire the operating parameters of the target HVAC;

[0138] An identification module 12, for performing fault identification on operating parameters using a preset fault identification model, and determining fault information of the target HVAC during operation;

[0139] Among them, the preset fault identification model is obtained by training the training results of the sample original operating parameters using the sample migration operating parameters under different migration scenarios; the sample migration operating parameters under different migration scenarios are based on the sample original operating parameters to predict the operating parameters corresponding to different operating conditions and different air-conditioning types.

[0140] In an exemplary embodiment, the identification module 12 includes a first identification unit, a second identification unit and a fault determination unit, wherein:

[0141] A first identification unit, configured to perform general fault identification on the operating parameters through a general fault model to obtain a first fault identification result;

[0142] A second identification unit is used to identify the migration fault of the operating parameter through the migration fault identification model to obtain a second fault identification result;

[0143] The fault determination unit is used to determine fault information of the target HVAC during operation based on the first fault identification result and the second fault identification result.

[0144] In an exemplary embodiment, the above-mentioned HVAC fault identification device further includes: a sample acquisition module, a first training module and a second training module, wherein:

[0145] A sample acquisition module is used to obtain sample original operation parameters of different types of HVAC during operation; and to obtain sample migration operation parameters under different migration scenarios; the sample original operation parameters and sample migration operation parameters both include normal operation parameters and fault operation parameters;

[0146] The first training module is used to train the initial fault recognition model through the original operating parameters of the sample to obtain an intermediate fault recognition model;

[0147] The second training module is used to train the intermediate fault identification model through sample migration operation parameters under different migration scenarios to obtain a fault identification model.

[0148] In an exemplary embodiment, the second training module comprises: a processing unit, a training unit and a model determination unit, wherein:

[0149] A processing unit, used for processing the intermediate fault identification model through different transfer learning methods to obtain multiple initial transfer fault identification models;

[0150] A training unit, used to train each initial migration fault identification model respectively by using sample migration operation parameters under different migration scenarios to obtain multiple migration fault identification models;

[0151] The model determination unit is used to determine the fault identification model based on each migration fault identification model.

[0152] In an exemplary embodiment, the model determination unit is further configured to evaluate multiple migration fault identification models respectively to obtain evaluation results of each migration fault identification model; and select the migration fault identification model with the best evaluation result from each evaluation result as the fault identification model.

[0153] In an exemplary embodiment, the above-mentioned model determination unit is also used to obtain performance evaluation parameters from the sample migration operation parameters under the different migration scenarios for any migration fault identification model; perform fault identification on the performance evaluation parameters through the migration fault identification model to obtain a first fault test result; and perform fault identification on the performance evaluation parameters through the intermediate fault identification model to obtain a second fault test result; determine the evaluation result corresponding to the migration fault identification model based on the first fault test result and the second fault test result.

[0154] In an exemplary embodiment, the above-mentioned model determination unit is also used to obtain a fault test error between the first fault test result and the second fault test result; the fault test error is evaluated based on a preset evaluation rule to obtain an evaluation result corresponding to the migration fault identification model.

[0155] Each module in the above-mentioned HVAC fault identification device can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.

[0156] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Fig.18As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data in the fault identification process of the HVAC. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for identifying faults of the HVAC is implemented.

[0157] Those skilled in the art will understand that Fig.18 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0158] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the content of any one embodiment of the above-mentioned HVAC fault identification method is implemented.

[0159] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the content of any one embodiment of the above-mentioned HVAC fault identification method is implemented.

[0160] In one embodiment, a computer program product is provided, including a computer program, which, when executed by a processor, implements the content of any one of the embodiments of the above-mentioned HVAC fault identification method.

[0161] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0162] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0163] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0164] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for identifying faults in a heating, ventilation and air conditioning system, characterized in that: The method comprises: Obtaining the operating parameters of the target HVAC; Performing fault identification on the operating parameters using a preset fault identification model to determine fault information of the target HVAC during operation; Among them, the preset fault identification model is obtained by training the training results of the sample original operating parameters using the sample migration operating parameters under different migration scenarios; the sample migration operating parameters under different migration scenarios are operating parameters corresponding to different operating conditions and different air-conditioning types predicted based on the sample original operating parameters.

2. The method according to claim 1, characterized in that The fault identification model includes a general fault model and a migration fault identification model; the method of using the preset fault identification model to perform fault identification on the operating parameters to determine the fault information of the target HVAC during operation includes: Performing universal fault identification on the operating parameters by using the universal fault model to obtain a first fault identification result; Identifying the migration fault of the operating parameter by using the migration fault identification model to obtain a second fault identification result; Based on the first fault identification result and the second fault identification result, fault information of the target HVAC during operation is determined.

3. The method according to claim 1 or 2, characterized in that: The construction process of the fault identification model includes: Obtaining sample original operating parameters of different types of HVAC during operation; and obtaining sample migration operating parameters under the different migration scenarios; the sample original operating parameters and the sample migration operating parameters both include normal operating parameters and fault operating parameters; The initial fault identification model is trained by using the original operating parameters of the sample to obtain an intermediate fault identification model; The intermediate fault identification model is trained by using the sample migration operation parameters under the different migration scenarios to obtain the fault identification model.

4. The method according to claim 3, characterized in that The training of the intermediate fault identification model by using the sample migration operation parameters in the different migration scenarios to obtain the fault identification model includes: Processing the intermediate fault identification model by different transfer learning methods to obtain multiple initial transfer fault identification models; The initial migration fault identification models are trained respectively by using the sample migration operation parameters under the different migration scenarios to obtain multiple migration fault identification models; The fault identification model is determined based on each of the migration fault identification models.

5. The method according to claim 4, characterized in that The determining the fault identification model based on each of the migration fault identification models includes: Evaluating the multiple migration fault identification models respectively to obtain evaluation results of the migration fault identification models; A migration fault identification model with the best evaluation result is selected from each of the evaluation results as the fault identification model.

6. The method according to claim 5, characterized in that The evaluating the multiple migration fault identification models respectively to obtain evaluation results of the migration fault identification models includes: For any migration fault identification model, obtaining performance evaluation parameters from sample migration operation parameters under the different migration scenarios; Performing fault identification on the performance evaluation parameter through the migration fault identification model to obtain a first fault test result; and performing fault identification on the performance evaluation parameter through the intermediate fault identification model to obtain a second fault test result; An evaluation result corresponding to the migration fault identification model is determined according to the first fault test result and the second fault test result.

7. The method according to claim 6, characterized in that The determining, according to the first fault test result and the second fault test result, an evaluation result corresponding to the migration fault identification model includes: Obtaining a fault test error between the first fault test result and the second fault test result; The fault test error is evaluated based on a preset evaluation rule to obtain an evaluation result corresponding to the migration fault identification model.

8. A fault identification device for HVAC, characterized in that: The device comprises: An acquisition module, used to acquire the operating parameters of the target HVAC; an identification module, configured to perform fault identification on the operating parameters using a preset fault identification model, and determine fault information of the target HVAC during operation; Among them, the preset fault identification model is obtained by training the training results of the sample original operating parameters using the sample migration operating parameters under different migration scenarios; the sample migration operating parameters under different migration scenarios are operating parameters corresponding to different operating conditions and different air-conditioning types predicted based on the sample original operating parameters.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images