Power supply abnormity diagnosis alarm method and system based on artificial intelligence

Abstract

The invention provides a power supply abnormity diagnosis and alarm method and system based on artificial intelligence, and the method comprises the steps: carrying out the goodness-of-fit matching of real-time power supply data and abnormal power supply data prepared in advance, and judging whether there is an abnormal condition or not. Through matching with the data corresponding to the abnormal condition with relatively high comprehensiveness, the abnormal early warning of the power supply data in a complex environment can be improved. When the goodness of fit between the real-time power supply data and the abnormal power supply data is reasoned, the characterization information of the data items in the abnormal power supply data, the characterization information of the data items in the environment parameters and the characterization information of the data items in the operation parameters are analyzed; considering the implied relationship between the characterization information of the data items in the abnormal power supply data and the characterization information of the data items in the environment parameters and the characterization information of the data items in the operation parameters in two dimensions of the operation parameters and the environment parameters so as to infer the goodness of fit between the real-time power supply data and the abnormal power supply data; and the reliability and the accuracy of the inference goodness of fit are improved.

Description

Technical Field

[0001] The present application relates to the field of data processing, and more specifically, to an artificial intelligence-based power supply abnormality diagnosis and alarm method and system. Background Art

[0002] The power supply is a key component of various computer systems and systems. It is responsible for storing electrical energy in other forms and converting it into electrical energy to supply the system when needed. Therefore, the stability and reliability of the power supply are crucial to the normal operation of the entire system. Power supply anomalies can manifest in many forms, such as unstable output voltage, no output or abnormal output voltage. These anomalies may be caused by a variety of reasons, including but not limited to excessive current, excessive voltage, aging of components, poor contact or equipment damage. Traditional power supply anomaly diagnosis methods mainly rely on manual inspection, empirical judgment and simple threshold comparison. These methods often fail to detect problems in a timely and accurate manner when faced with complex or novel power supply anomalies, resulting in low diagnostic efficiency or even misdiagnosis. With the development of artificial intelligence technology, anomaly detection methods based on machine learning and deep learning have been introduced into power supply anomaly diagnosis. These methods can identify anomalies by learning a large amount of historical data, improving the accuracy and efficiency of diagnosis. Modern power supply anomaly diagnosis methods have begun to comprehensively consider the inherent connections and mutual influences between multiple parameters and data streams. For example, in addition to considering the operating parameters of the power supply (such as voltage, current, etc.), environmental parameters (such as temperature, humidity, etc.) and the status information of the equipment are also combined for comprehensive judgment. This multi-parameter fusion diagnosis method helps to improve the accuracy and reliability of diagnosis. With the development of technologies such as the Internet of Things, big data, and cloud computing, power supply abnormality diagnosis is moving towards intelligence and automation. How to ensure the accuracy of power supply abnormality diagnosis is a technical problem that needs to be solved. Summary of the invention

[0003] The purpose of the present invention is to provide a power supply abnormality diagnosis and alarm method and system based on artificial intelligence. The technical solution of the embodiment of the present application is implemented as follows: In a first aspect, an embodiment of the present application provides a power supply abnormality diagnosis and alarm method based on artificial intelligence, the method comprising: Acquire a preset power supply abnormality data stream and a real-time power supply monitoring data stream, wherein the real-time power supply monitoring data stream includes an environmental parameter monitoring data stream and an operating parameter monitoring data stream of the real-time power supply data, wherein the preset power supply abnormality data stream represents data items in the abnormal power supply data, the environmental parameter monitoring data stream represents data items in the environmental parameters of the real-time power supply data, and the operating parameter monitoring data stream represents data items in the operating parameters of the real-time power supply data; Respectively extracting characterization information from the preset power supply abnormality data stream, the environmental parameter monitoring data stream, and the operating parameter monitoring data stream to obtain an abnormal power supply data characterization vector, an environmental parameter data characterization vector, and an operating parameter data characterization vector; Interacting the abnormal power data representation vector, the environmental parameter data representation vector, and the operating parameter data representation vector to obtain an interactive data representation vector; Performing anomaly diagnosis on the interactive data characterization vector to obtain a power supply anomaly diagnosis result, wherein the power supply anomaly diagnosis result indicates a degree of consistency between the real-time power supply data and the abnormal power supply data; When the power anomaly diagnosis result indicates that the consistency between the real-time power data and the abnormal power data reaches a preset requirement, a power anomaly alarm is issued.

[0004] As an implementation manner, the power supply abnormality diagnosis result is obtained based on a goodness of fit evaluation network, and the goodness of fit evaluation network includes an abnormal data embedding mapping layer, an environmental parameter embedding mapping layer, and an operating parameter embedding mapping layer; the characterization information of the preset power supply abnormality data stream, the environmental parameter monitoring data stream, and the operating parameter monitoring data stream are respectively extracted to obtain an abnormal power supply data characterization vector, an environmental parameter data characterization vector, and an operating parameter data characterization vector, including: According to the abnormal data embedding mapping layer, extracting the representation information of the preset power supply abnormal data stream to obtain the abnormal power supply data representation vector; According to the environmental parameter embedding mapping layer, extracting the representation information of the environmental parameter monitoring data stream to obtain the environmental parameter data representation vector; According to the operation parameter embedding mapping layer, the characterization information of the operation parameter monitoring data stream is extracted to obtain the operation parameter data characterization vector.

[0005] As an implementation manner, the extracting characterization information of the preset power supply abnormality data stream according to the abnormal data embedding mapping layer to obtain the abnormal power supply data characterization vector includes: Determine, according to the abnormal data embedding mapping layer, first power state characterization information and first distribution characterization information of the preset power abnormal data stream, wherein the first power state characterization information represents characterization information of multiple data items in the preset power abnormal data stream, and the first distribution characterization information represents characterization information of distribution conditions of multiple data items in the preset power abnormal data stream; Characterization information is extracted from the first power state characterization information and the first distribution characterization information to obtain the abnormal power data characterization vector.

[0006] As an implementation manner, extracting the representation information from the first power state representation information and the first distribution representation information to obtain the abnormal power data representation vector includes: Extracting representation information from the first power state representation information and the first distribution representation information to obtain a first anchor representation vector, a first result representation vector, and a first search representation vector; Interacting the first anchor representation vector, the first result representation vector, and the first search representation vector to obtain a candidate abnormal power data representation vector; Perform multiple rounds of nonlinear transformation on the candidate abnormal power supply data characterization vector to obtain the abnormal power supply data characterization vector.

[0007] As an implementation manner, the extracting of representation information from the environmental parameter monitoring data stream according to the environmental parameter embedding mapping layer to obtain the environmental parameter data representation vector includes: Determine, according to the environmental parameter embedding mapping layer, second power state characterization information and second distribution characterization information of the environmental parameter monitoring data stream, wherein the second power state characterization information represents characterization information of multiple data items in the environmental parameter monitoring data stream, and the second distribution characterization information represents characterization information of distribution of multiple data items in the environmental parameter monitoring data stream; Characterization information is extracted from the second power state characterization information and the second distribution characterization information to obtain the environmental parameter data characterization vector.

[0008] As an implementation manner, the extracting the representation information from the second power state representation information and the second distribution representation information to obtain the environmental parameter data representation vector includes: Extracting representation information from the second power state representation information and the second distribution representation information to obtain a second anchor representation vector, a second result representation vector, and a second search representation vector; Interacting the second anchor representation vector, the second result representation vector, and the second search representation vector to obtain a candidate environment parameter data representation vector; Perform multiple rounds of nonlinear transformation on the candidate environmental parameter data representation vector to obtain the environmental parameter data representation vector.

[0009] As an implementation manner, the extracting characterization information of the operating parameter monitoring data stream according to the operating parameter embedding mapping layer to obtain the operating parameter data characterization vector includes: According to the operating parameter embedding mapping layer, third power state characterization information and third distribution characterization information of the operating parameter monitoring data stream are determined, wherein the third power state characterization information represents characterization information of multiple data items in the operating parameter monitoring data stream, and the third distribution characterization information represents characterization information of the distribution of multiple data items in the operating parameter monitoring data stream; characterization information is extracted from the third power state characterization information and the third distribution characterization information to obtain the operating parameter data characterization vector.

[0010] As an implementation manner, the extracting the representation information from the third power state representation information and the third distribution representation information to obtain the operating parameter data representation vector includes: Extracting representation information from the third power state representation information and the third distribution representation information to obtain a third anchor representation vector, a third result representation vector, and a third search representation vector; Interacting the third anchor representation vector, the third result representation vector, and the third search representation vector to obtain a candidate operating parameter data representation vector; Perform multiple rounds of nonlinear transformation on the candidate operating parameter data representation vector to obtain the operating parameter data representation vector.

[0011] As an implementation mode, the consistency evaluation network is debugged by the following steps: Obtaining an example preset power supply abnormality data stream, an example real-time power supply monitoring data stream, and an actual power supply abnormality diagnosis result, wherein the example real-time power supply monitoring data stream includes an example environmental parameter monitoring data stream and an example operating parameter monitoring data stream, and the actual power supply abnormality diagnosis result indicates the actual consistency of the example real-time power supply data with the example abnormal power supply data; According to the consistency evaluation network, respectively extract characterization information of the example preset power supply abnormality data stream, the example environmental parameter monitoring data stream and the example operating parameter monitoring data stream to obtain an example abnormal power supply data characterization vector, an example environmental parameter data characterization vector and an example operating parameter data characterization vector; interact with the example abnormal power supply data characterization vector, the example environmental parameter data characterization vector and the example operating parameter data characterization vector to obtain an example interactive data characterization vector; perform abnormal diagnosis on the example interactive data characterization vector to obtain an example power supply abnormality diagnosis result; The consistency evaluation network is debugged according to the example power supply abnormality diagnosis result and the actual power supply abnormality diagnosis result.

[0012] As an implementation manner, the environment parameter embedding mapping layer is a debugged network layer, and the environment parameter embedding mapping layer is debugged by the following steps: Acquire a first environmental parameter monitoring data stream, shield data item information of a local area in the first environmental parameter monitoring data stream, and obtain a second environmental parameter monitoring data stream; Inferring the data item information of the shielded area in the second environmental parameter monitoring data stream according to the environmental parameter embedding mapping layer to obtain a second inference confidence, wherein the second inference confidence indicates the confidence that the data item of the shielded area in the first environmental parameter monitoring data stream obtained by inference belongs to each category of data items; Debugging the environmental parameter embedding mapping layer according to the second inference confidence and the second actual confidence, wherein the second actual confidence represents the actual confidence that the data items in the masked area in the first environmental parameter monitoring data stream belong to each category of data items; The operating parameter embedding mapping layer is a debugged network layer, and the operating parameter embedding mapping layer is debugged by the following steps: Acquire a first operating parameter monitoring data stream, shield data item information of a local area in the first operating parameter monitoring data stream, and obtain a second operating parameter monitoring data stream; Inferring the data item information of the masked area in the second operating parameter monitoring data stream according to the operating parameter embedding mapping layer to obtain a third inference confidence, wherein the third inference confidence indicates the confidence that the data item of the masked area in the first operating parameter monitoring data stream obtained by inference belongs to each category of data items; The operating parameter embedding mapping layer is debugged based on the third inference confidence and the third actual confidence, wherein the third actual confidence represents the actual confidence that the data items in the masked area in the first operating parameter monitoring data stream belong to each category of data items.

[0013] In a second aspect, the present application provides a computer system, comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor implements the method described above when executing the computer program.

[0014] Beneficial effects of the present application: The present application provides a power anomaly diagnosis and alarm method and system based on artificial intelligence, which matches the real-time power data with the abnormal power data prepared in advance to help monitor whether the power data has abnormal conditions. By matching with the data corresponding to the abnormal conditions with strong comprehensiveness, the early warning of abnormal power data in complex environments can be improved. When the degree of consistency between real-time power data and abnormal power data is inferred based on a neural network, the characterization information of the data items in the abnormal power data, the characterization information of the data items in the environmental parameters of the real-time power data, and the characterization information of the data items in the operating parameters of the real-time power data are analyzed, and the implicit relationship between the characterization information of the data items in the abnormal power data and the characterization information of the data items in the environmental parameters and the characterization information of the data items in the operating parameters is considered in the two dimensions of operating parameters and environmental parameters, so as to infer the degree of consistency between the real-time power data and the abnormal power data, thereby improving the reliability and accuracy of the inference degree of consistency.

[0015] Other features will be described in part in the following description. Those skilled in the art will discover these features in part upon inspection of the following content and drawings, or may learn these features through production or use. The features of the present application may be implemented and obtained by practicing or using various aspects of the methods, tools, and combinations listed in the detailed examples described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings herein are incorporated into the specification and constitute a part of the specification. These drawings illustrate embodiments consistent with the present application and are used together with the specification to illustrate the technical solution of the present application.

[0017] Figure 1 It is a schematic diagram of an application scenario of the artificial intelligence-based power anomaly diagnosis and alarm method provided in an embodiment of the present application.

[0018] Figure 2 This is a flow chart of an artificial intelligence-based power anomaly diagnosis and alarm method provided in an embodiment of the present application.

[0019] Figure 3 It is a schematic diagram of the composition of a computer system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0021] In the following description, references to “some embodiments”, “as an implementation / scheme”, and “in one implementation” describe a subset of all possible embodiments, but it can be understood that “some embodiments”, “as an implementation / scheme”, and “in one implementation” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0022] In the following description, the terms "first\second\third" and the like are only used to distinguish similar objects and do not represent a specific order for the objects. It is understandable that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present application described here can be implemented in an order other than that illustrated or described here. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application. The terms used herein are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.

[0023] The power anomaly diagnosis and alarm method based on artificial intelligence provided in the embodiment of the present application can be executed by a computer system, wherein the computer system can be various types of terminals such as a laptop, a tablet computer, a desktop computer, a mobile device (for example, a mobile phone, a portable music player, a personal digital assistant, a dedicated messaging device, a portable gaming device), etc., and can also be implemented as a server. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms.

[0024] Figure 1 Schematic diagram of the application scenario of the power anomaly diagnosis and alarm method based on artificial intelligence provided by the embodiment of the present application. It includes multiple sensors 100, a network 200 and a computer system 300, and the multiple sensors 100 and the computer system 300 are connected to each other through the network 200. The sensor 100 is used to detect the data of the target power supply and send the acquired data to the computer system 300 through the network 200. The computer system 300 is used to execute the method provided by the embodiment of the present application. Specifically, the embodiment of the present application provides a power anomaly diagnosis and alarm method based on artificial intelligence, such as Figure 2 As shown, the method includes: Step S110, obtaining a preset power abnormality data stream and a real-time power monitoring data stream, the real-time power monitoring data stream including an environmental parameter monitoring data stream and an operating parameter monitoring data stream of the real-time power data, the preset power abnormality data stream represents data items in the abnormal power data, the environmental parameter monitoring data stream represents data items in the environmental parameters of the real-time power data, and the operating parameter monitoring data stream represents data items in the operating parameters of the real-time power data.

[0025] In the embodiment of the present application, the preset power supply abnormality data stream is a set of predefined data, usually based on historical data or expert knowledge, used to represent the data characteristics that the power supply may exhibit in an abnormal state. These data streams contain various power supply data items that are considered abnormal. For example, in a data center environment, the preset power supply abnormality data stream may include a series of historical power supply temperature, voltage, and current values ​​when faults occurred, which are beyond the normal operating range and are therefore considered abnormal.

[0026] Real-time power monitoring data streams are data streams of power supply working status and environmental conditions collected in real time, which are used to reflect the current actual situation of the power supply. These data streams are constantly updated to provide the latest information about power supply performance. For example, the real-time working voltage, current, temperature and other data of a power supply are continuously collected and transmitted to the monitoring system to form a real-time power monitoring data stream.

[0027] The environmental parameter monitoring data stream is part of the real-time power supply monitoring data stream, which focuses on the state of the environment around the power supply, such as data on environmental factors such as temperature and humidity. These factors may affect the performance and life of the power supply. For example, the temperature sensor in the data center will continuously record the temperature of the computer room and send this data to the monitoring system to form an environmental parameter monitoring data stream. If the temperature exceeds the recommended operating range of the device, an alarm may be triggered.

[0028] The operating parameter monitoring data stream is also part of the real-time power monitoring data stream, focusing on the operating status of the power supply itself, including operating parameters such as voltage, current, and power factor. These data reflect the actual working efficiency and stability of the power supply. For example, the output voltage and current of the power supply will be monitored in real time and recorded as part of the operating parameter monitoring data stream. If the output voltage of the power supply fluctuates abnormally, this may indicate that there is a problem inside the power supply and further inspection is required.

[0029] A data item is a single, specific data element in a data stream. In the preset power anomaly data stream, environmental parameter monitoring data stream, and operating parameter monitoring data stream, each specific reading or measurement value is a data item.

[0030] Step S120 , respectively extracting characterization information from the preset power abnormality data stream, the environmental parameter monitoring data stream, and the operating parameter monitoring data stream to obtain an abnormal power data characterization vector, an environmental parameter data characterization vector, and an operating parameter data characterization vector.

[0031] After obtaining the preset power supply abnormality data stream, the environmental parameter monitoring data stream and the operating parameter monitoring data stream, the characterization information of the preset power supply abnormality data stream is extracted to obtain the abnormal power supply data characterization vector, which represents the type of each data item in the abnormal power supply data and the numerical values ​​of multiple data items. The characterization information of the environmental parameter monitoring data stream is extracted to obtain the environmental parameter data characterization vector, which represents the type of each data item in the environmental parameters of the real-time power supply data and the numerical values ​​of multiple data items. The characterization information of the operating parameter monitoring data stream is extracted to obtain the operating parameter data characterization vector, which represents the type of each data item in the operating parameters of the real-time power supply data and the numerical values ​​of multiple data items.

[0032] For example, for a preset power supply abnormality data stream, it contains records of power supply abnormalities that have occurred in the past. The computer system analyzes these data and extracts the features of the abnormal power supply data, such as abnormal voltage fluctuations, abnormal temperature rise, etc. These features will be converted into abnormal power supply data representation vectors. For example, an abnormal power supply data representation vector may contain a set of numbers, each number representing a specific abnormal feature, such as [0.8, 0.3, 0.5], where 0.8 may represent the degree of abnormal voltage fluctuation, 0.3 represents the degree of abnormal temperature rise, and 0.5 represents the degree of some other abnormal feature.

[0033] For the environmental parameter monitoring data stream, it reflects the real-time status of the environment where the power supply equipment is located, such as temperature, humidity, etc. The computer extracts key environmental parameters from these data streams, such as the current temperature, humidity value, etc., and converts them into environmental parameter data representation vectors. For example, an environmental parameter data representation vector may be in the form of [25, 60], where 25 represents the current ambient temperature (degrees Celsius) and 60 represents the current humidity percentage. For the operating parameter monitoring data stream, the real-time operating status of the power supply equipment is recorded, such as voltage, current, etc. The computer analyzes these data, extracts key operating parameters, such as input voltage, output current, etc., and converts them into operating parameter data representation vectors. For example, an operating parameter data representation vector may be in the form of [220, 5], where 220 represents the current input voltage (volts) and 5 represents the current output current (amperes).

[0034] Step S130 , interacting the abnormal power data representation vector, the environmental parameter data representation vector, and the operating parameter data representation vector to obtain an interactive data representation vector.

[0035] After obtaining the abnormal power data characterization vector, the environmental parameter data characterization vector, and the operating parameter data characterization vector, the abnormal power data characterization vector, the environmental parameter data characterization vector, and the operating parameter data characterization vector are interacted, that is, feature fusion is performed to obtain an interactive data characterization vector. The interactive data characterization vector contains feature information of abnormal power data, environmental parameters, and operating parameters, making the feature information of the interactive data characterization vector richer.

[0036] Step S140 , performing abnormal diagnosis on the interactive data characterization vector to obtain a power abnormality diagnosis result, where the power abnormality diagnosis result indicates the degree of consistency between the real-time power data and the abnormal power data.

[0037] Because the obtained interactive data characterization vector includes the characteristics of abnormal power data, environmental parameters and operating parameters, and the abnormal power data, environmental parameters and operating parameters contain the core information required for the matching degree of the characterization data, the abnormal power data and the power abnormality diagnosis results of the real-time power data can be obtained based on the abnormal diagnosis of the interactive data characterization vector. Among them, the process of abnormal diagnosis of the interactive data characterization vector is the process of performing the degree of matching reasoning based on the interactive data characterization vector. In other words, the abnormal power data and the power abnormality diagnosis results of the real-time power data are obtained by performing regression based on the interactive data characterization vector.

[0038] The above embodiments of the present application comprehensively consider the characterization information of data items in the abnormal power data, the characterization information of data items in the environmental parameters of the real-time power data, and the characterization information of data items in the operating parameters of the real-time power data when inferring the degree of consistency between the real-time power data and the abnormal power data. From the two dimensions of operating parameters and environmental parameters, the implicit connection between the characterization information of data items in the abnormal power data and the characterization information of data items in the environmental parameters and the characterization information of data items in the operating parameters is analyzed, thereby inferring the degree of consistency between the real-time power data and the abnormal power data. Not only are all the analysis dimensions complete, but the two dimensions of operating parameters and environmental parameters are used as matching directions, thereby increasing the accuracy of the inferred degree of consistency.

[0039] In one embodiment, the present application provides a goodness of fit evaluation network, which is used to infer the goodness of fit between real-time power supply data and abnormal power supply data. As a structural example of the goodness of fit evaluation network, the goodness of fit evaluation network may include: an abnormal data embedding mapping layer, an environmental parameter embedding mapping layer, an operating parameter embedding mapping layer, an interaction layer, and a dense layer.

[0040] Among them, the abnormal data embedding mapping layer, the environmental parameter embedding mapping layer, and the operating parameter embedding mapping layer are respectively connected to the interactive layer, and the interactive layer is connected to the dense layer. The abnormal data embedding mapping layer is used to extract the characteristic information of the preset power abnormal data stream, the environmental parameter embedding mapping layer is used to extract the characteristic information of the environmental parameter monitoring data stream, and the operating parameter embedding mapping layer is used to extract the characteristic information of the operating parameter monitoring data stream. The interactive layer is used to interact with the characteristic information extracted by the abnormal data embedding mapping layer, the environmental parameter embedding mapping layer, and the operating parameter embedding mapping layer (for example, fusion methods such as addition, splicing, and stacking), and the dense layer is used to perform abnormal diagnosis on the characteristic information to infer the power abnormal diagnosis result. The dense layer (Densely Connected layer) is also called a fully connected layer. The process of abnormal diagnosis and the process of full connection include feature integration, dimensionality reduction, or conversion to a more specific representation, so as to facilitate subsequent classification, regression or other tasks. In an embodiment of the present application, the dense layer is used to convert the interactive features into one or more numerical values, which represent the degree of fit of the real-time power data to the abnormal power data. If a specific fit value (regression task) is predicted, the output of the dense layer can be a continuous value representing the size of the fit. If the purpose is only to determine whether the inferred real-time power data and the abnormal power data have a high degree of consistency or a low degree of consistency (classification task), a softmax layer can be set after the dense layer to convert the output into a probability distribution for classification.

[0041] As an implementation design, the interaction layer may include a first filter, a second filter, and a third filter. The filter and the network component using the convolution kernel for filtering may be referred to as a convolution layer. The first filter is connected to the abnormal data embedding mapping layer, and the first filter is used to filter (i.e., convolve) the abnormal power data representation vector output by the abnormal data embedding mapping layer to obtain a deep abnormal power data representation vector; the second filter is connected to the environmental parameter embedding mapping layer, and the second filter is used to filter the environmental parameter data representation vector output by the environmental parameter embedding mapping layer to obtain a deep environmental parameter data representation vector; the third filter is connected to the operating parameter embedding mapping layer, and the third filter is used to filter the operating parameter data representation vector output by the operating parameter embedding mapping layer to obtain a deep operating parameter data representation vector.

[0042] For example, the execution data of the abnormal data embedded in the mapping layer is the preset power supply abnormality data stream, the execution data of the environmental parameter embedded in the mapping layer is the environmental parameter monitoring data stream, and the execution data of the operating parameter embedded in the mapping layer is the operating parameter monitoring data stream. The first filter, the second filter and the third filter respectively output the deep abnormal power supply data representation vector, the environmental parameter data representation vector and the operating parameter data representation vector obtained by filtering, and then the deep abnormal power supply data representation vector, the environmental parameter data representation vector and the operating parameter data representation vector are interacted to obtain the interactive data representation vector. The execution data of the dense layer is the interactive data representation vector, and the output of the dense layer is the power supply abnormality diagnosis result.

[0043] Step S150, when the power anomaly diagnosis result indicates that the consistency between the real-time power data and the abnormal power data reaches a preset requirement, a power anomaly alarm is issued.

[0044] It can be understood that the abnormal power data prepared in advance for the consistency comparison in the embodiment of the present application may include multiple. When a consistency comparison fails to meet the preset requirements (such as the consistency value is less than the preset value, or the display is inconsistent), the remaining abnormal power data and the real-time power data are sequentially determined to be consistent, and when the abnormal power data with the consistency meeting the preset requirements is determined, a power abnormality alarm is issued. The warning method is, for example, to execute the warning strategy corresponding to the abnormal power data that meets the preset requirements, such as broadcasting abnormal content, issuing a warning in a preset manner, and so on.

[0045] The above-mentioned consistency evaluation network can be used to infer the power supply abnormality diagnosis result. The consistency evaluation network includes: an abnormal data embedding mapping layer, an environmental parameter embedding mapping layer, an operating parameter embedding mapping layer, an interaction layer and a dense layer. Then, the method provided in the embodiment of the present application includes: Step S210, obtaining a preset power abnormality data stream and a real-time power monitoring data stream, the real-time power monitoring data stream including an environmental parameter monitoring data stream and an operating parameter monitoring data stream of the real-time power data, the preset power abnormality data stream represents data items in the abnormal power data, the environmental parameter monitoring data stream represents data items in the environmental parameters of the real-time power data, and the operating parameter monitoring data stream represents data items in the operating parameters of the real-time power data.

[0046] The process of step S210 may refer to the above-mentioned step S110.

[0047] Step S220 , based on the abnormal data embedded in the matching evaluation network into a mapping layer, characterization information is extracted from a preset power abnormality data stream to obtain an abnormal power data characterization vector.

[0048] The preset power supply abnormal data stream is input into the abnormal data embedding mapping layer in the consistency evaluation network, and the abnormal data embedding mapping layer extracts the representation information of the preset power supply abnormal data stream and outputs the abnormal power supply data representation vector.

[0049] As an implementation design, the network architecture of the abnormal data embedding mapping layer can be a convolutional neural network (CNN), a recurrent neural network (RNN) and its variants: such as a long short-term memory network (LSTM) or a gated recurrent unit (GRU). These network structures are suitable for processing sequence data and can capture the time dependency in the data. They are very suitable for processing power supply abnormality data streams, environmental parameter monitoring data streams and operating parameter monitoring data streams with timing characteristics.

[0050] As an implementation design, according to the abnormal data embedding mapping layer in the fit evaluation network, the characterization information of the preset power abnormal data stream is extracted to obtain the abnormal power data characterization vector, including: according to the abnormal data embedding mapping layer, the first power state characterization information and the first distribution characterization information of the preset power abnormal data stream are determined, and the characterization information of the first power state characterization information and the first distribution characterization information are extracted to obtain the abnormal power data characterization vector. The first power state characterization information represents the characterization information of multiple data items in the preset power abnormal data stream, and the first distribution characterization information represents the characterization information of the distribution of multiple data items in the preset power abnormal data stream.

[0051] The preset power abnormality data stream represents multiple data items in the abnormal power data. The preset power abnormality data stream may include data item features of multiple data items in the abnormal power data. In this application, a data item feature may be regarded as a token, and a token represents a data item. The first token in the preset power abnormality data stream is <start> ， <start>Represents the first token. The last token in the default power anomaly data stream is <end> ， <end>Represents the last token.

[0052] Based on the characteristics of multiple data items in the preset power abnormality data stream, the first power state characterization information and the first distribution characterization information can be determined, the first power state characterization information represents the characteristic information of the multiple data items in the preset power abnormality data stream, and the first distribution characterization information is the characterization information of the distribution of the multiple data items in the preset power abnormality data stream. According to the abnormal data embedding mapping layer, the characterization information of the first power state characterization information and the first distribution characterization information are extracted to obtain the abnormal power data characterization vector.

[0053] As an implementation manner, extracting the first power state representation information and the first distribution representation information to obtain the abnormal power data representation vector may include the following steps S221 to S223: Step S221, extracting representation information from the first power state representation information and the first distribution representation information to obtain a first anchor representation vector, a first result representation vector, and a first search representation vector.

[0054] The anchor representation vector is a key vector, the result representation vector is a value vector, and the search representation vector is a query vector. They belong to different vector domains. Each data item in the abnormal power supply data corresponds to its first anchor representation vector, first result representation vector, and first search representation vector. The first result representation vector represents the feature information of the data item itself, and the first search representation vector and the first anchor representation vector are used to determine the attention score.

[0055] As an implementation method, obtaining a first search representation vector, a first anchor representation vector, and a first result representation vector may specifically include: taking the product of the first power state representation information and the first parameter array as a semantic array, and obtaining the first search representation vector, the first anchor representation vector, and the first result representation vector corresponding to the first power state representation information according to the semantic array. The product of the first distribution representation information and the second parameter array is obtained to obtain a position array, and the first search representation vector, the first anchor representation vector, and the first result representation vector corresponding to the first distribution representation information are obtained according to the position array. The parameter array is usually a parameter in the form of a matrix. In the attention mechanism, a matrix representation of the search, key, and value is usually obtained from the input sequence through a linear transformation. These linear transformations are implemented by learnable parameter arrays (search array, key array, and value array). These parameter arrays are optimized during the training process so that the neural network can better capture the relevant information in the input data stream and generate effective outputs accordingly.

[0056] The first parameter array and the second parameter array are used for spatial mapping, and the first parameter array and the second parameter array are network parameters obtained through training. After obtaining the semantic array, the semantic array is split into a first search representation vector, a first anchor representation vector, and a first result representation vector corresponding to the first power state representation information. After obtaining the position array, the position array is split into a first search representation vector, a first anchor representation vector, and a first result representation vector corresponding to the first distribution representation information. For example, for the first parameter array, it is a parameter array with a dimension of three, and the semantic array obtained by multiplying the first power state representation information and the first parameter array is a semantic array with a dimension of three, and each dimension of the semantic array is respectively used as the first search representation vector, the first anchor representation vector, and the first result representation vector corresponding to the first power state representation information.

[0057] Step S222: Interact the first anchor characterization vector, the first result characterization vector, and the first search characterization vector to obtain a candidate abnormal power supply data characterization vector.

[0058] After obtaining the first anchor characterization vector, the first result characterization vector, and the first search characterization vector, the first anchor characterization vector, the first result characterization vector, and the first search characterization vector are interacted to obtain a candidate abnormal power supply data characterization vector.

[0059] As an implementation method, the first anchor characterization vector, the first result characterization vector and the first search characterization vector are interacted to obtain a candidate abnormal power data characterization vector, which may include, for example: standardizing the multiplication result of the first search characterization vector, the first anchor characterization vector and the normalization coefficient to obtain a standardized characterization vector, and using the multiplication result of the standardized characterization vector and the first result characterization vector as the candidate abnormal power data characterization vector. Obtain the normalization coefficient, wherein the normalization coefficient represents the standardized scaling ratio. The embodiment of the present application uses the normalization coefficient as a normalization parameter to standardize the multiplication result to obtain a standardized characterization vector, and the standardized characterization vector represents the correlation between the first search characterization vector and the first anchor characterization vector. Then, the standardized characterization vector can be used as the influence coefficient (i.e., weight) of the first result characterization vector, and the multiplication result between the standardized characterization vector and the first result characterization vector is determined as the candidate abnormal power data characterization vector.

[0060] The process of interacting the first anchor characterization vector, the first result characterization vector, and the first search characterization vector to obtain the candidate abnormal power data characterization vector can refer to the following formula: V1=f(S·M T ·α)·G; Specifically, V1 is the candidate abnormal power data representation vector, S is the first search representation vector, M is the first anchor representation vector, T is the transposition, α is the normalization coefficient, G is the first result representation vector, and f is the normalization function.

[0061] Step S223, performing multiple rounds of nonlinear transformation on the candidate abnormal power data characterization vector to obtain the abnormal power data characterization vector.

[0062] As an implementation method, multiple rounds of nonlinear transformations are performed on the candidate abnormal power data representation vector based on a multi-layer perceptron to obtain the abnormal power data representation vector. The multi-layer perceptron is represented as: V1'=MLP(V1).

[0063] Specifically, V1' is an abnormal power data representation vector, MLP is a multi-layer perceptron, and V1 is a candidate abnormal power data representation vector.

[0064] When extracting the abnormal power data characterization vector, for example, the input data is a token in a preset power abnormal data stream, for example, the token in the preset power abnormal data stream includes <start> 、<2>、<3>、<4>、… <end>. The power state characterization information and distribution characterization information of each data item are determined according to the token in the preset power abnormality data stream, and the power state characterization information and distribution characterization information of each data item are extracted according to the abnormal data embedding mapping layer to obtain the output, which is the abnormal power data characterization vector. The dimension of the abnormal power data characterization vector is consistent with the preset power abnormality data stream, and the abnormal power data characterization vector includes the deep features of each data item in the abnormal power data.

[0065] Step S230 , based on the environmental parameter embedding mapping layer in the fit evaluation network, characterization information is extracted from the environmental parameter monitoring data stream to obtain an environmental parameter data characterization vector.

[0066] The environmental parameter monitoring data stream is input into the environmental parameter embedding mapping layer in the consistency evaluation network, and the environmental parameter embedding mapping layer extracts the representation information of the environmental parameter monitoring data stream and outputs the environmental parameter data representation vector.

[0067] As an implementation design, the architecture of the environmental parameter embedding mapping layer can be a convolutional neural network (CNN), a recurrent neural network (RNN) and its variants: such as a long short-term memory network (LSTM) or a gated recurrent unit (GRU). These network structures are suitable for processing sequence data and can capture the temporal dependencies in the data. They are very suitable for processing power anomaly data streams, environmental parameter monitoring data streams, and operating parameter monitoring data streams with timing characteristics.

[0068] As an implementation design, based on the environmental parameter embedding mapping layer in the fit evaluation network, characterization information is extracted from the environmental parameter monitoring data stream to obtain an environmental parameter data characterization vector, which may specifically include: determining the second power state characterization information and the second distribution characterization information of the environmental parameter monitoring data stream based on the environmental parameter embedding mapping layer, extracting characterization information from the second power state characterization information and the second distribution characterization information to obtain an environmental parameter data characterization vector. The second power state characterization information represents the characterization information of multiple data items in the environmental parameter monitoring data stream, and the second distribution characterization information represents the characterization information of the distribution of multiple data items in the environmental parameter monitoring data stream.

[0069] The environmental parameter monitoring data stream represents multiple data items in the environmental parameter sequence. The environmental parameter monitoring data stream may include data item features of multiple data items in the environmental parameter data stream. As mentioned above, a data item feature may be regarded as a token, that is, a token represents a data item. The first token in the environmental parameter monitoring data stream is <start> ， <start>Represents the first token, and the last token in the environmental parameter monitoring data stream is <end> ， <end>Represents the last token. According to the characteristics of multiple data items in the environmental parameter monitoring data stream, the second power state characterization information and the second distribution characterization information can be determined. The second power state characterization information represents the characteristics of multiple data items in the environmental parameter monitoring data stream, and the second distribution characterization information represents the distribution of multiple data items in the environmental parameter monitoring data stream. According to the environmental parameter embedding mapping layer, the second power state characterization information and the second distribution characterization information are extracted to obtain the environmental parameter data characterization vector.

[0070] As an implementation manner, the method of extracting the representation information from the second power state representation information and the second distribution representation information to obtain the environmental parameter data representation vector may include the following steps S231 to S233: Step S231, extracting representation information from the second power state representation information and the second distribution representation information to obtain a second anchor representation vector, a second result representation vector and a second search representation vector.

[0071] Each data item in the environmental parameters includes a respective second anchor representation vector, a second result representation vector and a second search representation vector. The second result representation vector represents the characteristics of the data item itself, and the second search representation vector and the second anchor representation vector are used to determine the attention score.

[0072] As an implementation mode, obtaining the second search representation vector, the second anchor representation vector, and the second result representation vector includes: obtaining a semantic array by multiplying the second power state representation information with the third parameter array, and obtaining the second search representation vector, the second anchor representation vector, and the second result representation vector corresponding to the second power state representation information according to the semantic array. Obtaining a position array by multiplying the second distribution representation information with the fourth parameter array, and obtaining the second search representation vector, the second anchor representation vector, and the second result representation vector corresponding to the second distribution representation information according to the position array.

[0073] The third parameter array and the fourth parameter array are used for spatial mapping, and the third parameter array and the fourth parameter array are network parameters obtained through training.

[0074] Step S232: Interact the second anchor representation vector, the second result representation vector and the second search representation vector to obtain a candidate environment parameter data representation vector.

[0075] After obtaining the second anchor representation vector, the second result representation vector and the second search representation vector, the second anchor representation vector, the second result representation vector and the second search representation vector are interacted to obtain the candidate environment parameter data representation vector.

[0076] As an implementation method, the process of interacting the second anchor representation vector, the second result representation vector and the second search representation vector to obtain a candidate environmental parameter data representation vector includes: standardizing the multiplication result of the second search representation vector, the second anchor representation vector and the normalization coefficient to obtain a standardized representation vector, and determining the multiplication result of the standardized representation vector and the second result representation vector as the candidate environmental parameter data representation vector.

[0077] Step S233, performing multiple rounds of nonlinear transformation on the candidate environmental parameter data representation vector to obtain the environmental parameter data representation vector.

[0078] As an implementation method, a multi-layer perceptron is used to perform multiple rounds of nonlinear transformation on the candidate environmental parameter data representation vector to obtain the environmental parameter data representation vector.

[0079] Step S240 , based on the operating parameter embedding mapping layer in the fit evaluation network, characterization information is extracted from the operating parameter monitoring data stream to obtain an operating parameter data characterization vector.

[0080] The operating parameter monitoring data stream is input into the operating parameter embedding mapping layer in the goodness of fit evaluation network, and the operating parameter embedding mapping layer extracts the representation information of the operating parameter monitoring data stream and outputs the operating parameter data representation vector.

[0081] As an implementation design, the network architecture of the operating parameter embedding mapping layer can be a convolutional neural network (CNN), a recurrent neural network (RNN) and its variants: such as a long short-term memory network (LSTM) or a gated recurrent unit (GRU). These network structures are suitable for processing sequence data and can capture the time dependency in the data. They are very suitable for processing power anomaly data streams, environmental parameter monitoring data streams and operating parameter monitoring data streams with timing characteristics.

[0082] As an implementation design, based on the operating parameter embedding mapping layer in the fit evaluation network, characterization information is extracted from the operating parameter monitoring data stream to obtain an operating parameter data characterization vector, which may specifically include: determining the third power state characterization information and the third distribution characterization information of the operating parameter monitoring data stream based on the operating parameter embedding mapping layer, extracting characterization information from the third power state characterization information and the third distribution characterization information to obtain an operating parameter data characterization vector. The third power state characterization information represents characterization information of multiple data items in the operating parameter monitoring data stream, and the third distribution characterization information represents characterization information of the distribution of multiple data items in the operating parameter monitoring data stream. The operating parameter monitoring data stream represents multiple data items in the operating parameter sequence, and the operating parameter monitoring data stream may include data item features of multiple data items in the operating parameter data stream. As mentioned above, a data item feature can be regarded as a token, that is, a token represents a data item. The first token in the operating parameter monitoring data stream is <start> ， <start>Represents the first token, and the last token in the running parameter monitoring data stream is <end> ， <end>Represents the last token. According to the characteristics of multiple data items in the operating parameter monitoring data stream, the third power state characterization information and the third distribution characterization information can be determined. The third power state characterization information represents the characteristics of multiple data items in the operating parameter monitoring data stream, and the third distribution characterization information is the characterization information of the distribution of multiple data items in the operating parameter monitoring data stream. According to the operating parameter embedding mapping layer, the third power state characterization information and the third distribution characterization information are extracted to obtain the operating parameter data characterization vector.

[0083] As an implementation manner, extracting the representation information from the third power state representation information and the third distribution representation information to obtain the operation parameter data representation vector may specifically include the following steps S241 to S243: Step S241, extracting representation information from the third power state representation information and the third distribution representation information to obtain a third anchor representation vector, a third result representation vector, and a third search representation vector.

[0084] Each data item in the operating parameters includes a respective third anchor representation vector, a third result representation vector and a third search representation vector. The third result representation vector is a feature of the data item itself, and the third search representation vector and the third anchor representation vector are configured to determine an attention score.

[0085] As an implementation method, obtaining a third search representation vector, a third anchor representation vector, and a third result representation vector may specifically include: obtaining a semantic array by multiplying the third power state representation information with the fifth parameter array, and obtaining a third search representation vector, a third anchor representation vector, and a third result representation vector corresponding to the third power state representation information based on the semantic array. Obtain a position array by multiplying the third distribution representation information with the sixth parameter array, and obtain a third search representation vector, a third anchor representation vector, and a third result representation vector corresponding to the third distribution representation information based on the position array. The fifth parameter array and the sixth parameter array are used for spatial mapping, and the fifth parameter array and the sixth parameter array are network parameters obtained through training.

[0086] Step S242: interact with the third anchor representation vector, the third result representation vector and the third search representation vector to obtain a candidate operating parameter data representation vector.

[0087] After obtaining the third anchor characterization vector, the third result characterization vector and the third search characterization vector, the third anchor characterization vector, the third result characterization vector and the third search characterization vector are interacted to obtain the candidate operation parameter data characterization vector.

[0088] As an implementation method, the process of interacting the third anchor representation vector, the third result representation vector and the third search representation vector to obtain a candidate operating parameter data representation vector includes: standardizing the multiplication result of the third search representation vector, the third anchor representation vector and the normalization coefficient to obtain a standardized representation vector, and determining the multiplication result of the standardized representation vector and the third result representation vector as the candidate operating parameter data representation vector.

[0089] Step S243, performing multiple rounds of nonlinear transformation on the candidate operating parameter data representation vector to obtain the operating parameter data representation vector.

[0090] As an implementation method, a multi-layer perceptron is used to perform multiple rounds of nonlinear transformation on the candidate operating parameter data representation vector to obtain the operating parameter data representation vector.

[0091] Step S250 , based on the interaction layer in the fit evaluation network, the abnormal power data representation vector, the environmental parameter data representation vector and the operating parameter data representation vector are interacted to obtain an interaction data representation vector.

[0092] As an implementation design, the interaction layer includes a first filter, a second filter, and a third filter. The abnormal power data characterization vector is input into the first filter, and the first filter performs multiple filtering on the abnormal power data characterization vector to obtain a deep abnormal power data characterization vector. The environmental parameter data characterization vector is input into the second filter, and the environmental parameter data characterization vector is multiple filtered based on the second filter to obtain a deep environmental parameter data characterization vector. The operating parameter data characterization vector is input into the third filter, and the operating parameter data characterization vector is multiple filtered based on the third filter to obtain a deep operating parameter data characterization vector. The deep abnormal power data characterization vector, the environmental parameter data characterization vector, and the operating parameter data characterization vector are interacted (such as splicing) to obtain an interactive data characterization vector.

[0093] Step S260, performing abnormal diagnosis on the interactive data representation vector according to the dense layer in the fit evaluation network to obtain a power supply abnormality diagnosis result, wherein the power supply abnormality diagnosis result indicates the fit between the real-time power supply data and the abnormal power supply data.

[0094] The interactive data representation vector is input into the dense layer, and the dense layer performs abnormal diagnosis on the interactive data representation vector and outputs the power supply abnormality diagnosis result.

[0095] As an implementation design, the power anomaly diagnosis result is a consistency value. The larger the consistency value, the greater the consistency between the real-time power data and the abnormal power data. The smaller the consistency value, the smaller the consistency between the real-time power data and the abnormal power data. As mentioned above, in other implementation designs, the power anomaly diagnosis result is a binary classification result. When the power anomaly diagnosis result is the first result, the real-time power data is consistent with the abnormal power data. When the power anomaly diagnosis result is the second result, the real-time power data is inconsistent with the abnormal power data.

[0096] When inferring the degree of consistency between real-time power data and abnormal power data, the embodiment of the present application comprehensively measures the characterization information of data items in the abnormal power data, the characterization information of data items in the environmental parameters of the real-time power data, and the characterization information of data items in the operating parameters of the real-time power data. From the two aspects of operating parameters and environmental parameters, the implicit connection between the characterization information of data items in the abnormal power data and the characterization information of data items in the environmental parameters and the characterization information of data items in the operating parameters is considered to infer the degree of consistency between the real-time power data and the abnormal power data. The inferred degree of consistency has high accuracy and reliability.

[0097] The following describes the debugging process of the goodness of fit evaluation network, which can specifically include the following steps: Step S310, obtaining an example preset power anomaly data stream, an example real-time power monitoring data stream and an actual power anomaly diagnosis result, wherein the example real-time power monitoring data stream includes an example environmental parameter monitoring data stream and an example operating parameter monitoring data stream, and the actual power anomaly diagnosis result indicates the actual degree of fit between the example real-time power data and the example abnormal power data.

[0098] The example preset power supply abnormality data stream and the example real-time power supply monitoring data stream are training samples. The example real-time power supply monitoring data stream includes the example environmental parameter monitoring data stream and the example operating parameter monitoring data stream of the example real-time power supply data. The example preset power supply abnormality data stream represents data items in the example abnormal power supply data, the example environmental parameter monitoring data stream represents data items in the environmental parameters of the example real-time power supply data, and the example operating parameter monitoring data stream represents data items in the operating parameters of the example real-time power supply data.

[0099] The process of step S310 may refer to the aforementioned step S110.

[0100] Step S320, based on the consistency evaluation network, respectively extract characterization information of the example preset power supply abnormality data stream, the example environmental parameter monitoring data stream and the example operating parameter monitoring data stream to obtain the example abnormal power supply data characterization vector, the example environmental parameter data characterization vector and the example operating parameter data characterization vector.

[0101] The example preset power supply abnormality data stream, the example environmental parameter monitoring data stream and the example operating parameter monitoring data stream are input into the consistency evaluation network, and the consistency evaluation network extracts characterization information of the example preset power supply abnormality data stream, the example environmental parameter monitoring data stream and the example operating parameter monitoring data stream, and outputs the example abnormal power supply data characterization vector, the example environmental parameter data characterization vector and the example operating parameter data characterization vector.

[0102] As an implementation design, the goodness of fit evaluation network includes an abnormal data embedding mapping layer, an environmental parameter embedding mapping layer, and an operating parameter embedding mapping layer. Obtaining an example abnormal power data characterization vector, an example environmental parameter data characterization vector, and an example operating parameter data characterization vector may specifically include the following steps S321 to S323.

[0103] Step S321 , extracting characterization information from an example preset power supply abnormality data stream according to the abnormal data embedding mapping layer, and obtaining an example abnormal power supply data characterization vector.

[0104] As an implementation mode, according to the abnormal data embedding mapping layer, the first example power state characterization information and the first example distribution characterization information of the example preset power abnormal data stream are determined, and the characterization information of the first example power state characterization information and the first example distribution characterization information are extracted to obtain the example abnormal power data characterization vector. The first example power state characterization information represents the characterization information of multiple data items in the example preset power abnormal data stream, and the first example distribution characterization information represents the characterization information of the distribution of multiple data items in the example preset power abnormal data stream.

[0105] As an implementation mode, extracting the representation information of the first example power state representation information and the first example distribution representation information to obtain the example abnormal power data representation vector includes: extracting the representation information of the first example power state representation information and the first example distribution representation information to obtain the first example anchor representation vector, the first example result representation vector and the first example search representation vector. Interacting the first example anchor representation vector, the first example result representation vector and the first example search representation vector to obtain the candidate example abnormal power data representation vector. Performing multiple rounds of nonlinear transformation on the candidate example abnormal power data representation vector to obtain the example abnormal power data representation vector.

[0106] Step S322: extracting characterization information from the example environmental parameter monitoring data stream according to the environmental parameter embedding mapping layer to obtain a characterization vector of the example environmental parameter data.

[0107] As an implementation mode, according to the environmental parameter embedding mapping layer, the second example power state characterization information and the second example distribution characterization information of the example environmental parameter monitoring data stream are determined, and the characterization information of the second example power state characterization information and the second example distribution characterization information are extracted to obtain the example environmental parameter data characterization vector. The second example power state characterization information represents the characterization information of multiple data items in the example environmental parameter monitoring data stream, and the second example distribution characterization information represents the characterization information of the distribution of multiple data items in the example environmental parameter monitoring data stream.

[0108] As an implementation, the process of extracting representation information from the second example power state representation information and the second example distribution representation information to obtain the example environment parameter data representation vector includes: extracting representation information from the second example power state representation information and the second example distribution representation information to obtain the second example anchor representation vector, the second example result representation vector, and the second example search representation vector. Interacting the second example anchor representation vector, the second example result representation vector, and the second example search representation vector to obtain the candidate example environment parameter data representation vector. Performing multiple rounds of nonlinear transformation on the candidate example environment parameter data representation vector to obtain the example environment parameter data representation vector.

[0109] Step S323: extracting characterization information from the example operating parameter monitoring data stream according to the operating parameter embedding mapping layer to obtain a characterization vector of the example operating parameter data.

[0110] As an implementation mode, according to the operation parameter embedding mapping layer, the third example power state characterization information and the third example distribution characterization information of the example operation parameter monitoring data stream are determined, and the characterization information is extracted from the third example power state characterization information and the third example distribution characterization information to obtain the example operation parameter data characterization vector. The third example power state characterization information represents the characterization information of multiple data items in the example operation parameter monitoring data stream, and the third example distribution characterization information represents the characterization information of the distribution of multiple data items in the example operation parameter monitoring data stream.

[0111] As an implementation manner, the process of extracting representation information from the third example power state representation information and the third example distribution representation information to obtain the example operation parameter data representation vector includes: extracting representation information from the third example power state representation information and the third example distribution representation information to obtain the third example anchor representation vector, the third example result representation vector, and the third example search representation vector. Interacting the third example anchor representation vector, the third example result representation vector, and the third example search representation vector to obtain the example candidate operation parameter data representation vector. Performing multiple rounds of nonlinear transformation on the example candidate operation parameter data representation vector to obtain the example operation parameter data representation vector.

[0112] Step S330 , based on the goodness of fit evaluation network, interact with the example abnormal power data representation vector, the example environmental parameter data representation vector and the example operating parameter data representation vector to obtain an example interactive data representation vector.

[0113] As an implementation design, the goodness of fit evaluation network includes an interaction layer, based on which an example abnormal power data representation vector, an example environmental parameter data representation vector, and an example operating parameter data representation vector are interacted to obtain an example interaction data representation vector.

[0114] As an implementation method, the interaction layer includes a first filter, a second filter, and a third filter. An example abnormal power data characterization vector is input into the first filter, and the example abnormal power data characterization vector is filtered multiple times based on the first filter to obtain a deep example abnormal power data characterization vector. An example environmental parameter data characterization vector is input into the second filter, and the example environmental parameter data characterization vector is filtered multiple times based on the second filter to obtain a deep example environmental parameter data characterization vector. An example operation parameter data characterization vector is input into the third filter, and the example operation parameter data characterization vector is filtered multiple times based on the third filter to obtain a deep example operation parameter data characterization vector. The deep example abnormal power data characterization vector, the example environmental parameter data characterization vector, and the example operation parameter data characterization vector are interacted to obtain an example interaction data characterization vector.

[0115] Step S340 , performing an abnormality diagnosis on the example interaction data representation vector according to the goodness of fit evaluation network to obtain an example power supply abnormality diagnosis result.

[0116] As an implementation design, the goodness of fit evaluation network includes a dense layer, based on which an abnormality diagnosis is performed on the example interaction data representation vector to obtain an example power supply abnormality diagnosis result.

[0117] Step S350 , debugging the consistency evaluation network according to the example power supply abnormality diagnosis result and the actual power supply abnormality diagnosis result.

[0118] The actual power anomaly diagnosis result of the example abnormal power data and the example real-time power data is obtained. The actual power anomaly diagnosis result indicates the actual degree of consistency of the example real-time power data with the example abnormal power data. The obtained example power anomaly diagnosis result is based on the degree of consistency inferred by the consistency evaluation network. According to the difference between the example power anomaly diagnosis result and the actual power anomaly diagnosis result, the consistency evaluation network is debugged.

[0119] Because the goal of the fit evaluation network is to infer the actual fit of the example real-time power data to the example abnormal power data, the more the example power abnormal diagnosis result matches the actual power abnormal diagnosis result, the more accurate the fit evaluation network is. According to the error between the example power abnormal diagnosis result and the actual power abnormal diagnosis result, the fit evaluation network is debugged to reduce the error between the example power abnormal diagnosis result obtained by the debugged fit evaluation network and the actual power abnormal diagnosis result, increase the reasoning effect of the fit evaluation network, and improve the accuracy of the fit evaluation network.

[0120] As an implementation design, the above steps S310 to S350 are repeated to complete the iterative debugging of the fit evaluation network until a preset debugging cutoff condition is reached, such as the number of debugging reaches a preset maximum number or the network loss is less than a preset loss value.

[0121] When debugging a consistency evaluation network for inferring the consistency between real-time power data and abnormal power data, the embodiment of the present application comprehensively measures the characterization information of data items in the abnormal power data, the characterization information of data items in the environmental parameters of the real-time power data, and the characterization information of data items in the operating parameters of the real-time power data. From the two dimensions of operating parameters and environmental parameters, the consistency evaluation network learns the implicit relationship between the characterization information of data items in the abnormal power data and the characterization information of data items in the environmental parameters, and the characterization information of data items in the operating parameters, and infers the consistency between the real-time power data and the abnormal power data. Not only all dimensions are considered, but also analysis is performed from the two perspectives of operating parameters and environmental parameters, so that the reasoning effect of the consistency evaluation network is improved.

[0122] The above-mentioned goodness of fit evaluation network includes an abnormal data embedding mapping layer, an environmental parameter embedding mapping layer, and an operating parameter embedding mapping layer. Before debugging the goodness of fit evaluation network, the abnormal data embedding mapping layer, the environmental parameter embedding mapping layer, and the operating parameter embedding mapping layer are untrained embedding mapping layers, or are embedded mapping layers after pre-training. After the pre-training is completed, the goodness of fit evaluation network is established based on the abnormal data embedding mapping layer, the environmental parameter embedding mapping layer, and the operating parameter embedding mapping layer obtained by the pre-training, and then the goodness of fit evaluation network is debugged.

[0123] The pre-debugging process of embedding abnormal data into the mapping layer, embedding environmental parameters into the mapping layer, and embedding operating parameters into the mapping layer is as follows.

[0124] Step S410, obtaining a first preset power abnormality data stream, shielding data item information of a local area in the first preset power abnormality data stream, and obtaining a second preset power abnormality data stream.

[0125] The first preset power abnormality data stream is a preset power abnormality data stream of any abnormal power data, and the first preset power abnormality data stream represents a data item in the abnormal power data, and the abnormal power data includes multiple data items. The first preset power abnormality data stream includes data item information of multiple regions, and the data item of each region represents the data item of the corresponding region. The data item information of the local region in the first preset power abnormality data stream is arbitrarily shielded to obtain the second preset power abnormality data stream, and the data item of the local region in the second preset power abnormality data stream is an unknown data item.

[0126] As an implementation design, the data items in the local area of ​​the first preset power supply abnormality data stream are set as preset marks, for example, the preset mark is <h>.

[0127] Step S420, based on the abnormal data embedding mapping layer, infer the data item information of the masked area in the second preset power supply abnormality data stream to obtain a first inference confidence, and the first inference confidence represents the confidence that the data items in the masked area in the first preset power supply abnormality data stream obtained by inference belong to each category of data items.

[0128] The second preset power supply abnormality data stream is input into the abnormal data embedding mapping layer, and the abnormal data embedding mapping layer infers the data item information of the masked area in the second preset power supply abnormality data stream, and outputs a first inference confidence.

[0129] As an implementation design, during pre-debugging, compared with the consistency evaluation network, the abnormal data embedding mapping layer also includes an output structure. After the pre-debugging of the abnormal data embedding mapping layer is completed, the output structure in the abnormal data embedding mapping layer is deleted, and a consistency evaluation network is established based on the abnormal data embedding mapping layer after deleting the output structure. According to the abnormal data embedding mapping layer, the power state characterization information and distribution characterization information of the second preset power abnormality data stream are determined, and the characterization information of the power state characterization information and the distribution characterization information are extracted to obtain the abnormal power data characterization vector. Among them, the power state characterization information represents the characterization information of multiple data items in the preset power abnormality data stream, and the distribution characterization information represents the characterization information of the distribution of multiple data items in the preset power abnormality data stream. Then, the output structure is used to infer the abnormal power data characterization vector to obtain the first inference confidence.

[0130] The manner in which the abnormal data embedding mapping layer extracts the abnormal power data characterization vector of the second preset power abnormality data stream is consistent with the manner in which the abnormal power data characterization vector is extracted in the above embodiment.

[0131] Step S430, debugging the abnormal data embedding mapping layer according to the first inference confidence and the first actual confidence, the first actual confidence representing the actual confidence that the data items in the masked area in the first preset power abnormality data stream belong to each type of data items.

[0132] A first actual confidence is obtained, where the first actual confidence represents the actual confidence that the data items in the masked area in the first preset power supply abnormality data stream belong to each category of data items. The obtained first inference confidence is inferred by the abnormal data embedding mapping layer. According to the difference between the first inference confidence and the first actual confidence, the abnormal data embedding mapping layer is debugged.

[0133] Because the goal of the abnormal data embedding mapping layer is to infer the actual confidence that the data items in the masked area in the first preset power supply abnormal data stream belong to each type of data item, the more similar the first inference confidence is to the first actual confidence, the better the effect of the abnormal data embedding mapping layer in extracting the abnormal power supply data representation vector. According to the error between the first inference confidence and the first actual confidence, the first inference confidence and the first actual confidence are debugged to reduce the error between the first inference confidence obtained according to the debugged abnormal data embedding mapping layer and the first actual confidence, thereby increasing the representation information extraction capability of the abnormal data embedding mapping layer, so that the abnormal data embedding mapping layer can roughly learn to extract the abnormal power supply data representation vector.

[0134] In the above implementation mode, the abnormal data embedding mapping layer is pre-debugged based on self-supervision according to the example preset power abnormality data stream, which increases the effect of the abnormal data embedding mapping layer on extracting characterization information of the preset power abnormality data stream. Subsequently, a consistency evaluation network is established based on the pre-debugged abnormal data embedding mapping layer, which helps to reduce the debugging complexity of the consistency evaluation network and ensure the accuracy and reliability of the consistency evaluation network.

[0135] The pre-debugging process of embedding environment parameters into the mapping layer may include: Step S510, obtaining a first environmental parameter monitoring data stream, shielding data item information of a local area in the first environmental parameter monitoring data stream, and obtaining a second environmental parameter monitoring data stream.

[0136] The first environmental parameter monitoring data stream can be an environmental parameter monitoring data stream of any type of environmental parameter. The first environmental parameter monitoring data stream represents data items in the environmental parameters, and the environmental parameters include multiple data items. The first environmental parameter monitoring data stream includes data item information of multiple regions, and the data item information of each region represents the data items of the region. The data item information of the local region in the first environmental parameter monitoring data stream is arbitrarily shielded to obtain the second environmental parameter monitoring data stream, and then the data item information of the local region in the second environmental parameter monitoring data stream is unknown. The shielding method of the first environmental parameter monitoring data stream can refer to the shielding method of the aforementioned first preset power supply abnormality data stream.

[0137] Step S520, based on the environmental parameter embedded mapping layer, infer the data item information of the shielded area in the second environmental parameter monitoring data stream to obtain a second inference confidence, and the second inference confidence represents the confidence that the data items in the shielded area in the first environmental parameter monitoring data stream obtained by inference belong to each category of data items.

[0138] The second environmental parameter monitoring data stream is input into the environmental parameter embedding mapping layer, and the environmental parameter embedding mapping layer infers the data item information of the shielded area in the second environmental parameter monitoring data stream and outputs a second inference confidence.

[0139] As an implementation design, during pre-debugging, compared with the consistency evaluation network, the environmental parameter embedding mapping layer also includes an output structure. After the pre-debugging of the environmental parameter embedding mapping layer is completed, the output structure in the environmental parameter embedding mapping layer is deleted, and a consistency evaluation network is established based on the environmental parameter embedding mapping layer after deleting the output structure. Then, based on the environmental parameter embedding mapping layer, the power state characterization information and distribution characterization information of the second environmental parameter monitoring data stream are determined, and the characterization information of the power state characterization information and the distribution characterization information are extracted to obtain the environmental parameter data characterization vector. Among them, the power state characterization information represents the characterization information of multiple data items in the environmental parameter monitoring data stream, and the distribution characterization information represents the characterization information of the distribution of multiple data items in the environmental parameter monitoring data stream. Then, the environmental parameter data characterization vector is inferred based on the output structure to obtain a second inference confidence. The method for extracting the environmental parameter data characterization vector of the second environmental parameter monitoring data stream by the environmental parameter embedding mapping layer can refer to the method for extracting the environmental parameter data characterization vector in the aforementioned embodiment.

[0140] Step S530, based on the second inference confidence and the second actual confidence, the debugging environment parameter embedding mapping layer, the second actual confidence represents the actual confidence that the data items in the masked area of ​​the first environment parameter monitoring data stream belong to each category of data items.

[0141] A second actual confidence is obtained, where the second actual confidence represents the actual confidence that the data items in the masked area in the first environmental parameter monitoring data stream belong to each category of data items. The obtained second inference confidence is inferred based on the environmental parameter embedding mapping layer. The environmental parameter embedding mapping layer is debugged based on the error between the second inference confidence and the second actual confidence.

[0142] Because the goal of the environmental parameter embedding mapping layer is to infer the actual confidence that the data items in the masked area in the first environmental parameter monitoring data stream belong to each type of data item, the more similar the second inference confidence is to the second actual confidence, the better the effect of the environmental parameter embedding mapping layer in extracting the environmental parameter data representation vector. Then, according to the error between the second inference confidence and the second actual confidence, the second inference confidence and the second actual confidence are debugged to reduce the error between the second inference confidence obtained according to the debugged environmental parameter embedding mapping layer and the second actual confidence, so as to increase the representation information extraction effect of the environmental parameter embedding mapping layer, thereby making the environmental parameter embedding mapping layer familiar with extracting the environmental parameter data representation vector.

[0143] The embodiment of the present application performs pre-debugging of the environmental parameter embedding mapping layer through a self-supervision mechanism based on the example environmental parameter monitoring data stream, thereby increasing the effect of the environmental parameter embedding mapping layer in extracting characterization information on the environmental parameter monitoring data stream, and then establishes a consistency evaluation network based on the pre-debugging environmental parameter embedding mapping layer, thereby helping to reduce the debugging complexity of the consistency evaluation network and ensuring the reliability and accuracy of the consistency evaluation network.

[0144] The following describes the pre-debugging process of embedding environment parameters into the mapping layer, which includes the following steps: Step S610, obtaining a first operating parameter monitoring data stream, shielding data item information of a local area in the first operating parameter monitoring data stream, and obtaining a second operating parameter monitoring data stream.

[0145] The first operating parameter monitoring data stream can be an operating parameter monitoring data stream of any type of operating parameter, and the first operating parameter monitoring data stream represents the data items in the operating parameters, and the operating parameters include multiple data items. The first operating parameter monitoring data stream includes data item information of multiple regions, and the data item information of each region represents the data items of the region. The data item information of the local region in the first operating parameter monitoring data stream is arbitrarily shielded to obtain the second operating parameter monitoring data stream, and then the data item information of the local region in the second operating parameter monitoring data stream is unknown. The shielding method of the first operating parameter monitoring data stream can refer to the shielding method of the first preset power supply abnormality data stream in the above embodiment.

[0146] Step S620, based on the operating parameter embedded mapping layer, infer the data item information of the masked area in the second operating parameter monitoring data stream to obtain a third inference confidence, and the third inference confidence represents the confidence that the data items in the masked area in the first operating parameter monitoring data stream obtained by inference belong to each category of data items.

[0147] The second operating parameter monitoring data stream is loaded into the operating parameter embedding mapping layer, and the data item information of the shielded area in the second operating parameter monitoring data stream is inferred based on the operating parameter embedding mapping layer, and the third inference confidence is output. As an implementation design, during the pre-debugging, compared with the consistency evaluation network, the operating parameter embedding mapping layer also includes an output structure. After the pre-debugging of the operating parameter embedding mapping layer is completed, the output structure in the operating parameter embedding mapping layer is deleted, and the consistency evaluation network is constructed based on the operating parameter embedding mapping layer after the output structure is deleted. Then, according to the operating parameter embedding mapping layer, the power state characterization information and distribution characterization information of the second operating parameter monitoring data stream are determined, and the characterization information of the power state characterization information and the distribution characterization information are extracted to obtain the operating parameter data characterization vector. Among them, the power state characterization information represents the characterization information of multiple data items in the operating parameter monitoring data stream, and the distribution characterization information represents the characterization information of the distribution of multiple data items in the operating parameter monitoring data stream. Then, the operating parameter data characterization vector is inferred based on the output structure to obtain the third inference confidence.

[0148] The method for extracting the operating parameter data characterization vector of the second operating parameter monitoring data stream by the operating parameter embedding mapping layer may refer to the above-mentioned method for extracting the operating parameter data characterization vector.

[0149] Step S630, based on the third inference confidence and the third actual confidence, debug the operation parameter embedding mapping layer, the third actual confidence represents the actual confidence that the data items in the masked area in the first operation parameter monitoring data stream belong to each category of data items.

[0150] A third actual confidence is obtained, where the third actual confidence represents the actual confidence that the data items in the masked area in the first operating parameter monitoring data stream belong to each category of data items. The obtained third inference confidence is inferred based on the operating parameter embedding mapping layer. According to the error between the third inference confidence and the third actual confidence, the operating parameter embedding mapping layer is debugged.

[0151] Because the goal of the operation parameter embedding mapping layer is to infer the actual confidence that the data items in the masked area in the first operation parameter monitoring data stream belong to each type of data items, the more similar the third inference confidence is to the third actual confidence, the better the effect of the operation parameter embedding mapping layer in extracting the operation parameter data representation vector. Then, based on the error between the third inference confidence and the third actual confidence, the third inference confidence and the third actual confidence are debugged to reduce the error between the third inference confidence obtained according to the debugged operation parameter embedding mapping layer and the third actual confidence, thereby increasing the representation information extraction effect of the operation parameter embedding mapping layer, so that the operation parameter embedding mapping layer is familiar with the process of extracting the operation parameter data representation vector.

[0152] The present application performs pre-debugging of the operating parameter embedding mapping layer through a self-supervision mechanism based on an example operating parameter monitoring data stream, thereby increasing the effect of the operating parameter embedding mapping layer in extracting characterization information on the operating parameter monitoring data stream, and then establishes a fit evaluation network based on the pre-debugging operating parameter embedding mapping layer, thereby helping to reduce the debugging complexity of the fit evaluation network and ensuring the accuracy and reliability of the fit evaluation network.

[0153] It should be noted that in the embodiment of the present application, if the above method is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the embodiment of the present application is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, which is stored in a storage medium, including a number of instructions to enable a computer system (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM), a disk or an optical disk. In this way, the embodiment of the present application is not limited to any specific combination of hardware and software.

[0154] An embodiment of the present application provides a computer system, including a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the above method when executing the computer program.

[0155] The embodiment of the present application provides a computer-readable storage medium on which a computer program is stored, and the computer program implements the above method when executed by a processor. The computer-readable storage medium may be transient or non-transient.

[0156] The embodiment of the present application provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and when the computer program is read and executed by a computer, some or all of the steps in the above method are implemented. The computer program product can be implemented specifically by hardware, software or a combination thereof. In an optional embodiment, the computer program product is embodied as a computer storage medium, and in another optional embodiment, the computer program product is embodied as a software product, such as a software development kit (Software Development Kit, SDK) and the like.

[0157] It should be noted that Figure 3 is a hardware entity diagram of a computer system 300 provided in an embodiment of the present application, such as Figure 3 As shown, the hardware entity of the computer system 300 includes: a processor 310, a communication interface 320 and a memory 330, wherein: the processor 310 generally controls the overall operation of the computer system 300. The communication interface 320 enables the computer system to communicate with other terminals or servers through a network. The memory 330 is configured to store instructions and applications executable by the processor 310, and can also cache data to be processed or processed by the processor 310 and each module in the computer system 300 (for example, image data, audio data, voice communication data and video communication data), which can be implemented by flash memory (FLASH) or random access memory (Random Access Memory, RAM). Data can be transmitted between the processor 310, the communication interface 320 and the memory 330 through the bus 340. It should be pointed out here that the description of the above storage medium and device embodiments is similar to the description of the above method embodiments, and has similar beneficial effects as the method embodiments. For technical details not disclosed in the storage medium and device embodiments of the present application, please refer to the description of the method embodiments of the present application for understanding.

[0158] It should be understood that "one embodiment" or "an embodiment" mentioned throughout the specification means that specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present application, the size of the serial number of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application. The above-mentioned serial numbers of the embodiments of the present application are for description only and do not represent the advantages and disadvantages of the embodiments.

[0159] It should be noted that, in this article, the term "includes", "comprising" or any other variation thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0160] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0161] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0162] In addition, all functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may be separately configured as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0163] A person skilled in the art can understand that all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM), magnetic disks or optical disks, etc., various media that can store program codes.

[0164] Alternatively, if the above-mentioned integrated unit of the present application is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can essentially or in other words, the part that contributes to the relevant technology can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for a computer system (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROMs, magnetic disks, or optical disks.

[0165] The above is only an implementation method of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.< / h> < / end> < / end> < / start> < / start> < / end> < / end> < / start> < / start> < / end> < / start> < / end> < / end> < / start> < / start>

Claims

1. A power supply abnormality diagnosis and alarm method based on artificial intelligence, characterized in that: The method comprises: Acquire a preset power supply abnormality data stream and a real-time power supply monitoring data stream, wherein the real-time power supply monitoring data stream includes an environmental parameter monitoring data stream and an operating parameter monitoring data stream of the real-time power supply data, wherein the preset power supply abnormality data stream represents data items in the abnormal power supply data, the environmental parameter monitoring data stream represents data items in the environmental parameters of the real-time power supply data, and the operating parameter monitoring data stream represents data items in the operating parameters of the real-time power supply data; Respectively extracting characterization information from the preset power supply abnormality data stream, the environmental parameter monitoring data stream, and the operating parameter monitoring data stream to obtain an abnormal power supply data characterization vector, an environmental parameter data characterization vector, and an operating parameter data characterization vector; Interacting the abnormal power data representation vector, the environmental parameter data representation vector, and the operating parameter data representation vector to obtain an interactive data representation vector; Performing anomaly diagnosis on the interactive data characterization vector to obtain a power supply anomaly diagnosis result, wherein the power supply anomaly diagnosis result indicates a degree of consistency between the real-time power supply data and the abnormal power supply data; When the power anomaly diagnosis result indicates that the consistency between the real-time power data and the abnormal power data reaches a preset requirement, a power anomaly alarm is issued.

2. The method according to claim 1, characterized in that The power supply abnormality diagnosis result is obtained based on a goodness of fit evaluation network, wherein the goodness of fit evaluation network includes an abnormal data embedding mapping layer, an environmental parameter embedding mapping layer, and an operating parameter embedding mapping layer; the characterization information of the preset power supply abnormality data stream, the environmental parameter monitoring data stream, and the operating parameter monitoring data stream are extracted respectively to obtain an abnormal power supply data characterization vector, an environmental parameter data characterization vector, and an operating parameter data characterization vector, including: According to the abnormal data embedding mapping layer, extracting the representation information of the preset power supply abnormal data stream to obtain the abnormal power supply data representation vector; According to the environmental parameter embedding mapping layer, extracting the representation information of the environmental parameter monitoring data stream to obtain the environmental parameter data representation vector; According to the operation parameter embedding mapping layer, the characterization information of the operation parameter monitoring data stream is extracted to obtain the operation parameter data characterization vector.

3. The method according to claim 2, characterized in that The extracting characterization information of the preset power supply abnormality data stream according to the abnormal data embedding mapping layer to obtain the abnormal power supply data characterization vector includes: Determine, according to the abnormal data embedding mapping layer, first power state characterization information and first distribution characterization information of the preset power abnormal data stream, wherein the first power state characterization information represents characterization information of multiple data items in the preset power abnormal data stream, and the first distribution characterization information represents characterization information of distribution conditions of multiple data items in the preset power abnormal data stream; Characterization information is extracted from the first power state characterization information and the first distribution characterization information to obtain the abnormal power data characterization vector.

4. The method according to claim 3, characterized in that The extracting the representation information from the first power state representation information and the first distribution representation information to obtain the abnormal power data representation vector includes: Extracting representation information from the first power state representation information and the first distribution representation information to obtain a first anchor representation vector, a first result representation vector, and a first search representation vector; Interacting the first anchor representation vector, the first result representation vector, and the first search representation vector to obtain a candidate abnormal power data representation vector; Perform multiple rounds of nonlinear transformation on the candidate abnormal power supply data characterization vector to obtain the abnormal power supply data characterization vector.

5. The method according to claim 2, characterized in that: The step of extracting characterization information from the environmental parameter monitoring data stream according to the environmental parameter embedding mapping layer to obtain the environmental parameter data characterization vector includes: Determine, according to the environmental parameter embedding mapping layer, second power state characterization information and second distribution characterization information of the environmental parameter monitoring data stream, wherein the second power state characterization information represents characterization information of multiple data items in the environmental parameter monitoring data stream, and the second distribution characterization information represents characterization information of distribution of multiple data items in the environmental parameter monitoring data stream; Characterization information is extracted from the second power state characterization information and the second distribution characterization information to obtain the environmental parameter data characterization vector.

6. The method according to claim 5, characterized in that The extracting the representation information from the second power state representation information and the second distribution representation information to obtain the environmental parameter data representation vector includes: Extracting representation information from the second power state representation information and the second distribution representation information to obtain a second anchor representation vector, a second result representation vector, and a second search representation vector; Interacting the second anchor representation vector, the second result representation vector, and the second search representation vector to obtain a candidate environment parameter data representation vector; Perform multiple rounds of nonlinear transformation on the candidate environmental parameter data representation vector to obtain the environmental parameter data representation vector.

7. The method according to claim 2, characterized in that The extracting characterization information of the operating parameter monitoring data stream according to the operating parameter embedding mapping layer to obtain the operating parameter data characterization vector includes: According to the operating parameter embedding mapping layer, third power state characterization information and third distribution characterization information of the operating parameter monitoring data stream are determined, wherein the third power state characterization information represents characterization information of multiple data items in the operating parameter monitoring data stream, and the third distribution characterization information represents characterization information of the distribution of multiple data items in the operating parameter monitoring data stream; characterization information is extracted from the third power state characterization information and the third distribution characterization information to obtain the operating parameter data characterization vector.

8. The method according to claim 7, characterized in that The extracting the representation information from the third power state representation information and the third distribution representation information to obtain the operating parameter data representation vector includes: Extracting representation information from the third power state representation information and the third distribution representation information to obtain a third anchor representation vector, a third result representation vector, and a third search representation vector; Interacting the third anchor representation vector, the third result representation vector, and the third search representation vector to obtain a candidate operating parameter data representation vector; Perform multiple rounds of nonlinear transformation on the candidate operating parameter data representation vector to obtain the operating parameter data representation vector.

9. The method according to any one of claims 2 to 8, characterized in that: The consistency evaluation network is debugged by the following steps: Obtaining an example preset power supply abnormality data stream, an example real-time power supply monitoring data stream, and an actual power supply abnormality diagnosis result, wherein the example real-time power supply monitoring data stream includes an example environmental parameter monitoring data stream and an example operating parameter monitoring data stream, and the actual power supply abnormality diagnosis result indicates the actual consistency of the example real-time power supply data with the example abnormal power supply data; According to the consistency evaluation network, respectively extract characterization information of the example preset power supply abnormality data stream, the example environmental parameter monitoring data stream and the example operating parameter monitoring data stream to obtain an example abnormal power supply data characterization vector, an example environmental parameter data characterization vector and an example operating parameter data characterization vector; interact with the example abnormal power supply data characterization vector, the example environmental parameter data characterization vector and the example operating parameter data characterization vector to obtain an example interactive data characterization vector; perform abnormal diagnosis on the example interactive data characterization vector to obtain an example power supply abnormality diagnosis result; The consistency evaluation network is debugged according to the example power supply abnormality diagnosis result and the actual power supply abnormality diagnosis result.

10. The method according to claim 9, characterized in that The environment parameter embedding mapping layer is a debugged network layer, and the environment parameter embedding mapping layer is debugged by the following steps: Acquire a first environmental parameter monitoring data stream, shield data item information of a local area in the first environmental parameter monitoring data stream, and obtain a second environmental parameter monitoring data stream; Inferring the data item information of the shielded area in the second environmental parameter monitoring data stream according to the environmental parameter embedding mapping layer to obtain a second inference confidence, wherein the second inference confidence indicates the confidence that the data item of the shielded area in the first environmental parameter monitoring data stream obtained by inference belongs to each category of data items; Debugging the environmental parameter embedding mapping layer according to the second inference confidence and the second actual confidence, wherein the second actual confidence represents the actual confidence that the data items in the masked area in the first environmental parameter monitoring data stream belong to each category of data items; The operating parameter embedding mapping layer is a debugged network layer, and the operating parameter embedding mapping layer is debugged by the following steps: Acquire a first operating parameter monitoring data stream, shield data item information of a local area in the first operating parameter monitoring data stream, and obtain a second operating parameter monitoring data stream; Inferring the data item information of the masked area in the second operating parameter monitoring data stream according to the operating parameter embedding mapping layer to obtain a third inference confidence, wherein the third inference confidence indicates the confidence that the data item of the masked area in the first operating parameter monitoring data stream obtained by inference belongs to each category of data items; The operating parameter embedding mapping layer is debugged based on the third inference confidence and the third actual confidence, wherein the third actual confidence represents the actual confidence that the data items in the masked area in the first operating parameter monitoring data stream belong to each category of data items.

11. A computer system comprising a memory and a processor, characterized in that: The memory stores a computer program that can be run on the processor, and the processor implements the method according to any one of claims 1 to 10 when executing the computer program.

Images