A multi-index simultaneous anomaly diagnosis method and system based on reconstruction probability

Abstract

The application provides a multi-index simultaneous anomaly diagnosis method and system based on reconstruction probability, and relates to the technical field of data processing. The method comprises the following steps: inputting the obtained multi-index continuous sampling data sequence of a certain monitoring station into the encoding network structure of a variational autoencoder to obtain the first reconstruction probability of the mean and standard deviation distribution of each index data input; inputting normal index data into the encoding network structure of the variational autoencoder to obtain the second reconstruction probability of the distribution of the corresponding index of the normal data; comparing the size of the first reconstruction probability and the second reconstruction probability, if the first reconstruction probability is less than the second reconstruction probability, and the product of the first reconstruction probability and the second reconstruction probability is less than a preset probability threshold, it is considered that the multiple indexes of the certain monitoring station have occurred abnormal; otherwise, it is considered that the multiple indexes of the certain monitoring station have not occurred abnormal. The application simultaneously judges the abnormal conditions of multiple indexes, and improves the accuracy and reliability of water quality anomaly detection.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a diagnostic method and system for simultaneous anomalies of multiple indicators based on reconstruction probability. Background Technology

[0002] When surface water pollution incidents occur, multiple indicators (two or more) usually show abnormalities simultaneously. However, if only a single indicator is monitored for anomalies, the abnormality may not be obvious, leading to missed reports. Therefore, it is necessary to assess the anomalies of multiple indicators simultaneously.

[0003] Therefore, the urgent technical problem to be solved is: how to provide a diagnostic method and system for multiple indicators that are simultaneously abnormal based on reconstruction probability, so as to judge the abnormality of multiple indicators at the same time and improve the accuracy and reliability of water quality anomaly detection. Summary of the Invention

[0004] The purpose of this application is to provide a diagnostic method and system for simultaneous anomalies of multiple indicators based on reconstruction probability, which can simultaneously judge the abnormality of multiple indicators and improve the accuracy and reliability of water quality anomaly detection.

[0005] To achieve the above objectives, as a first aspect of this application, this application provides a diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability. The method includes: acquiring a continuous sampling data sequence of multiple indicators from a monitoring station; inputting the acquired continuous sampling data sequence of multiple indicators from the monitoring station into the encoding network structure of a variational autoencoder to obtain a first reconstruction probability of the mean and standard deviation distribution of each input indicator data; inputting normal indicator data into the encoding network structure of the variational autoencoder to obtain a second reconstruction probability of the distribution of the corresponding indicators of the normal data; comparing the magnitudes of the first reconstruction probability and the second reconstruction probability; if the first reconstruction probability is less than the second reconstruction probability, and the product of the first reconstruction probability and the second reconstruction probability is less than a preset probability threshold, then it is considered that multiple indicators of the monitoring station are abnormal; otherwise, it is considered that multiple indicators of the monitoring station are not abnormal.

[0006] The diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability, as described above, further includes: if multiple indicators of a monitoring station are abnormal, calculating the comprehensive anomaly score of the multiple indicators.

[0007] The diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability, as described above, wherein the continuous sampling data sequence of multiple indicators is represented as: A = {A1, A2, A3, ..., A...} M};A M Represents a continuous data sequence of the Mth indicator; A1 = {A 11 A 12 A13 A 1N};A 1N This represents the sampled data of the first indicator at time N.

[0008] The diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability, as described above, wherein the coding network structure of the variational autoencoder includes an encoder and a decoder.

[0009] The diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability, as described above, wherein the encoder adopts a convolution + LSTM structure, and the final output layer unit is the number of indicators * 2.

[0010] As described above, the diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability involves the decoder sampling the mean and standard deviation of each indicator, and then fusing the outputs of each layer into the same dimension as the input data.

[0011] As a second aspect of this application, this application provides a diagnostic system for simultaneous anomalies of multiple indicators based on reconstruction probability. The system includes: a data acquisition module for acquiring a continuous sampling data sequence of multiple indicators from a monitoring station; a network module for inputting the acquired continuous sampling data sequence of multiple indicators from the monitoring station into the encoding network structure of a variational autoencoder to obtain a first reconstruction probability of the mean and standard deviation distribution of each input indicator data; the network module is further used to input normal indicator data into the encoding network structure of the variational autoencoder to obtain a second reconstruction probability of the distribution of the corresponding indicators of the normal data; and a data comparator for comparing the magnitudes of the first reconstruction probability and the second reconstruction probability. If the first reconstruction probability is less than the second reconstruction probability, and the product of the first reconstruction probability and the second reconstruction probability is less than a preset probability threshold, then multiple indicators of the monitoring station are considered to be abnormal; otherwise, multiple indicators of the monitoring station are considered not to be abnormal.

[0012] The diagnostic system for simultaneous anomalies of multiple indicators based on reconstruction probability, as described above, further includes a data processor for calculating a comprehensive anomaly score for multiple indicators if multiple indicators of a monitoring station are abnormal.

[0013] The diagnostic system for simultaneous anomalies of multiple indicators based on reconstruction probability, as described above, includes a variational autoencoder encoding network structure in the network module; the variational autoencoder encoding network structure includes an encoder and a decoder.

[0014] The diagnostic system for simultaneous anomalies of multiple indicators based on reconstruction probability, as described above, wherein the encoder adopts a convolutional + LSTM structure.

[0015] The beneficial effects achieved by this application are as follows:

[0016] (1) This application optimizes the model structure of traditional VAE, which can diagnose anomalies of multiple indicators and multiple time points at the same time, and can process multiple indicators at the same time, thereby improving the accuracy and reliability of water quality anomaly detection.

[0017] (2) The long short-term memory network used in this application has the following advantages:

[0018] Capturing long-term dependencies: Through its unique design structure, particularly the introduced "memory units," LSTM is able to effectively learn long-term dependencies in sequential data. This ability allows LSTM to accurately predict important events when processing time series data, even with very long intervals and delays, which is difficult to achieve in traditional RNNs.

[0019] Addressing the vanishing or exploding gradient problem: Standard RNNs are prone to gradient vanishing or exploding when processing long sequences of data, making the model difficult to train. LSTM overcomes this limitation by introducing a "gating" mechanism (forget gate, input gate, and output gate) to finely control the reading, writing, and output of information, enabling the model to effectively learn long-term dependencies. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings.

[0021] Figure 1 This is a flowchart illustrating a diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability, according to an embodiment of this application.

[0022] Figure 2 This is a schematic diagram of the structure of a diagnostic system for simultaneous anomalies of multiple indicators based on reconstruction probability, according to an embodiment of this application. Detailed Implementation

[0023] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0024] Example 1

[0025] like Figure 1 As shown, this application provides a diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability. The method includes the following steps:

[0026] Step S1: Obtain a multi-indicator continuous sampling data sequence for a monitoring station.

[0027] Among them, the multi-indicator continuous sampling data sequence is: the data of M indicators at N consecutive time points of a certain monitoring station.

[0028] The multi-index continuous sampling data sequence is represented as: A = {A1, A2, A3, ..., A...} M};A M This represents a continuous data sequence for the Mth indicator.

[0029] Where, A1={A 11 A 12 A 13 A 1N}. A 1N This represents the sampled data of the first indicator at time N.

[0030] Step S2: Input the multi-index continuous sampling data sequence of a certain monitoring station into the coding network structure of the variational autoencoder to obtain the first reconstruction probability of the mean and standard deviation distribution of each input index data.

[0031] The mean of each indicator is represented as: μ = {μ1, μ2, μ3, ..., μM}. μM represents the mean of the Mth indicator. The standard deviation of each indicator is: a = {a1, a2, a3, ..., aM}. aM represents the standard deviation of the Mth indicator.

[0032] The formula for calculating the first reconstruction probability is as follows:

[0033]

[0034] Where CP represents the first reconstruction probability; U N U1 represents the index data value at time N within the sampling time; U2 represents the index data value at time 1 within the sampling time. denoted by ; μ represents the mean of the index data; and a represents the standard deviation of the index data.

[0035] The coding network structure of a variational autoencoder includes an encoder and a decoder.

[0036] The encoder employs a convolutional + LSTM (Long Short-Term Memory) structure, combining convolutional and LSTM structures. The final output layer unit has twice the number of metrics. This differs from traditional VAEs (Variational Autoencoders). The use of a long short-term memory network in this application offers the following advantages:

[0037] Capturing long-term dependencies: Through its unique design structure, particularly the introduced "memory units," LSTM is able to effectively learn long-term dependencies in sequential data. This ability allows LSTM to accurately predict important events when processing time series data, even with very long intervals and delays, which is difficult to achieve in traditional RNNs.

[0038] Addressing the vanishing or exploding gradient problem: Standard RNNs are prone to gradient vanishing or exploding when processing long sequences of data, making the model difficult to train. LSTM overcomes this limitation by introducing a "gating" mechanism (forget gate, input gate, and output gate) to finely control the reading, writing, and output of information, enabling the model to effectively learn long-term dependencies.

[0039] LSMT stands for Long Short-Term Memory Network, which is used to solve the problem of long-term dependencies, that is, to remember information over a long time interval when processing sequential data.

[0040] The basic structure of LSTM consists of four main components: input gate, forget gate, output gate, and memory cell (also known as cell state).

[0041] The input gate is denoted as i. t The _i_{i} ... t =σ(W i ·[h t-1 x t ]+b i );

[0042] Among them W i and W C These are the weight matrices for the input gate and the candidate cell state, respectively. i and b C This is the corresponding bias term. t This is the input for the current time step. This is the candidate cell state.

[0043] The Forget Gate is denoted as f. t The sigmoid function determines which information in a cellular state should be forgotten. It also uses the sigmoid function to determine the importance of past memories; a value closer to 1 indicates that most information is retained, and a value closer to 0 indicates that most information is forgotten. The mathematical expression is: f t =σ(W f ·[h t-1 x t ]+b f );

[0044] Where σ is the sigmoid activation function, W f This is the weight matrix of the forget gate, b f It is the bias term of the forget gate, h t-1 It is the hidden state of the previous time step, x t This is the input for the current time step.

[0045] Cell state is denoted as C t The is the core of LSTM, a vector capable of storing long-term information. It is updated by combining a dot product operation with the forget gate and the cell state from the previous time step, along with the input gate and a new candidate memory state. The candidate memory state is obtained by applying the tanh activation function to the current input and its weight matrix. The mathematical expression is:

[0046] Candidate cell state, denoted as At each time step, the current input x t and the previous hidden state h t-1 The information is processed to generate a candidate cell state, which contains information that may be added to the long-term state.

[0047] The output gate is denoted as o. t The `tf.h` function controls which information from the cell state should be used to generate the output of the current time step. It combines the `sigmoid` function (which determines which cell state information should be output) and the `tanh` function (which scales the selected memory to ensure the output is between -1 and 1). The mathematical expression is: o t =σ(W o ·[h t-1 x t ]+b o );

[0048] h t =o t *tanh(C t );

[0049] Among them W o It is the weight matrix of the output gate, b o This is the bias term for the output gate. t-1 It is the hidden state of the previous time step, x t This is the input for the current time step.

[0050] The decoder samples the mean and standard deviation of each indicator and then merges the outputs of each layer into the same dimension as the input data.

[0051] Step S3: Input the normal index data into the coding network structure of the variational autoencoder to obtain the second reconstruction probability of the distribution of the corresponding index of the normal data.

[0052] The second reconstruction probability of the distribution of corresponding indicators for normal data can be evaluated using an autoencoder for anomaly detection and state monitoring. An autoencoder is an artificial neural network that learns efficient encodings of data values ​​in an unsupervised manner. The purpose of this network structure is to learn an representation (encoding) of a set of data, often used for dimensionality reduction. Both the decoder and encoder learn, enabling the autoencoder to attempt to generate an expression from the dimensionality-reduced encoding that is as close as possible to the original input. In the context of anomaly detection and state monitoring, the basic idea is to use an autoencoder to "compress" sensor readings into a lower-dimensional representation that captures the correlations and interactions between various variables. By training the autoencoder network using data representing a "normal" operating state, the goal is to be able to reconstruct the input variables after dimensionality reduction. The main idea is that as the performance of the monitored equipment degrades, this affects the interactions between variables. When the equipment state changes, the error in reconstructing the input variables by the neural network increases significantly. By monitoring the reconstruction error, an indication of the "health status" of the monitored equipment can be obtained, as the reconstruction error increases with equipment degradation.

[0053] The calculation method for the second reconstruction probability is the same as that for the first reconstruction probability.

[0054] Step S4: Compare the magnitudes of the first reconstruction probability and the second reconstruction probability. If the first reconstruction probability is less than the second reconstruction probability, and the product of the first reconstruction probability and the second reconstruction probability is less than a preset probability threshold, then it is considered that multiple indicators of a certain monitoring station have become abnormal; otherwise, it is considered that multiple indicators of a certain monitoring station have not become abnormal.

[0055] Step S5: If multiple indicators of a monitoring station are abnormal, calculate the comprehensive abnormality score of the multiple indicators.

[0056] Specifically, the formula for calculating the comprehensive abnormality score of multiple indicators is as follows:

[0057]

[0058] Where Zy represents the composite anomaly score of multiple indicators; M represents the number of indicators; B β (Xi,N) represents the encoding result of the data at time N in the i-th index data sequence input into the encoder; B β (Z) represents the encoding result of the normal data of the i-th indicator input into the encoder; β represents the encoder parameters; Xi represents the input data sequence of the i-th indicator, and N represents the N-th time step; J γ[Xi,N)] represents the result obtained by inputting the data at time N in the i-th index data sequence into the decoder; J γ (Zi) represents the result of normal data input to the decoder for the i-th indicator; γ represents the parameters of the decoder.

[0059] Example 2

[0060] like Figure 2 As shown, this application provides a diagnostic system 100 for simultaneous anomalies of multiple indicators based on reconstruction probability. The system includes:

[0061] The data acquisition module 10 is used to acquire a multi-indicator continuous sampling data sequence of a monitoring station.

[0062] Network module 20 is used to input the acquired multi-index continuous sampling data sequence of a certain monitoring station into the coding network structure of the variational autoencoder to obtain the first reconstruction probability of the mean and standard deviation distribution of each input index data.

[0063] Network module 20 is also used to input normal index data into the coding network structure of the variational autoencoder to obtain the second reconstruction probability of the distribution of the corresponding index of the normal data.

[0064] The data comparator 30 is used to compare the magnitude of the first reconstruction probability and the second reconstruction probability. If the first reconstruction probability is less than the second reconstruction probability, and the product of the first reconstruction probability and the second reconstruction probability is less than a preset probability threshold, then it is considered that multiple indicators of a certain monitoring station have become abnormal; otherwise, it is considered that multiple indicators of a certain monitoring station have not become abnormal.

[0065] Data processor 40 is used to calculate the comprehensive abnormality score of multiple indicators if multiple indicators of a monitoring station are abnormal.

[0066] Network module 20 sets up the encoding network structure of the variational autoencoder. The encoding network structure of the variational autoencoder includes an encoder and a decoder. The encoder uses a convolutional + LSTM structure.

[0067] This application also provides a computer storage medium storing computer instructions, which, when invoked, execute the address mapping method of the large-capacity solid-state drive. The computer storage medium includes one or more program instructions, which are executed by a processor to provide a diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability.

[0068] The embodiments disclosed in this invention provide a computer-readable storage medium storing computer program instructions. When the computer program instructions are executed on a computer, the computer performs the aforementioned diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability.

[0069] This invention provides a processor for processing the above-described diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability.

[0070] In this embodiment of the invention, the processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0071] The various methods, steps, and logic diagrams disclosed in the embodiments of this invention can be implemented or executed. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The processor reads information from the storage medium and, in conjunction with its hardware, completes the steps of the above methods.

[0072] The storage medium can be memory, such as volatile memory or non-volatile memory, or may include both volatile and non-volatile memory.

[0073] The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EEPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DRRAM).

[0074] The beneficial effects achieved by this application are as follows:

[0075] (1) This application optimizes the model structure of traditional VAE, which can diagnose anomalies of multiple indicators and multiple time points at the same time, and can process multiple indicators at the same time, thereby improving the accuracy and reliability of water quality anomaly detection.

[0076] (2) The long short-term memory network used in this application has the following advantages:

[0077] Capturing long-term dependencies: Through its unique design structure, particularly the introduced "memory units," LSTM is able to effectively learn long-term dependencies in sequential data. This ability allows LSTM to accurately predict important events when processing time series data, even with very long intervals and delays, which is difficult to achieve in traditional RNNs.

[0078] Addressing the vanishing or exploding gradient problem: Standard RNNs are prone to gradient vanishing or exploding when processing long sequences of data, making the model difficult to train. LSTM overcomes this limitation by introducing a "gating" mechanism (forget gate, input gate, and output gate) to finely control the reading, writing, and output of information, enabling the model to effectively learn long-term dependencies.

[0079] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0080] In the description of this application, the word "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0081] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability, characterized in that, This method is applied to the detection of water quality anomalies, and the method includes: Obtain a continuous sampling data sequence of multiple indicators for a certain monitoring station; where the continuous sampling data sequence of multiple indicators is represented as: A = {A1, A2, A3, ..., A...} M };A M This represents a continuous data sequence for the Mth indicator; A1={ A 11 A 12 A 13 A 1N };A 1N This represents the sampled data of the first indicator at time N; The acquired multi-index continuous sampling data sequence from a monitoring station is input into the coding network structure of a variational autoencoder to obtain the first reconstruction probability of the mean and standard deviation distribution of each input index data; wherein, the formula for calculating the first reconstruction probability is: ； ； in, Indicates the first reconstruction probability; This represents the index data value at time N within the sampling period; This represents the index data value at time 1 within the sampling period; Represents the cumulative distribution function; This represents the mean of the indicator data; This represents the standard deviation of the indicator data; Normal indicator data is input into the encoding network structure of a variational autoencoder to obtain the second reconstruction probability of the distribution of the corresponding indicators of normal data. The second reconstruction probability is evaluated by using an autoencoder for anomaly detection and state monitoring. In the context of anomaly detection and state monitoring, the autoencoder is used to compress sensor readings into a lower-dimensional representation, capturing the correlation and interaction between various variables. The autoencoder network is trained using data representing normal operating conditions. The input variables are reconstructed after dimensionality reduction. By monitoring the reconstruction error, the health status indication of the monitored device is obtained. Compare the magnitudes of the first reconstruction probability and the second reconstruction probability. If the first reconstruction probability is less than the second reconstruction probability, and the product of the first reconstruction probability and the second reconstruction probability is less than a preset probability threshold, then multiple indicators of a certain monitoring station are considered to be abnormal; otherwise, multiple indicators of a certain monitoring station are considered not to be abnormal. The coding network structure of the variational autoencoder includes an encoder and a decoder; the encoder adopts a convolutional + LSTM structure, and the final output layer unit is the number of indices * 2. If multiple indicators at a monitoring station are abnormal, calculate the combined abnormality score of the multiple indicators; The formula for calculating the comprehensive abnormal score of multiple indicators is as follows: ； in, This represents the combined abnormal score of multiple indicators; Indicates the number of indicators; Indicates the first The data at time N in a sequence of index data is input into the encoder as the encoded result. Indicates the first The normal data of each indicator is input into the encoder to obtain the encoded result. Indicates the parameters of the encoder; Indicates the input number There are several indicator data sequences, where N represents the Nth time point; Indicates the first The result obtained by inputting the data at time N in a sequence of index data into the decoder; Indicates the first The normal data of each indicator is input into the decoder as a result; This represents the parameters of the decoder.

2. The diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability according to claim 1, characterized in that, The mean of each indicator is represented as: μ={μ1,μ2,μ3,…,μM}; μM represents the mean of the Mth indicator data; The standard deviation of each indicator data is: a = {a1, a2, a3, ..., aM}; aM represents the standard deviation of the Mth indicator data.

3. The diagnostic method for simultaneous anomalies of multiple indicators based on reconstruction probability according to claim 1, characterized in that, The decoder samples the mean and standard deviation of each indicator, and then merges the outputs of each layer into the same dimension as the input data.

4. A diagnostic system for simultaneous anomalies of multiple indicators based on reconstruction probability, characterized in that, The system for performing the method according to any one of claims 1-3 includes: The data acquisition module is used to acquire continuous sampling data sequences of multiple indicators at a monitoring station; The network module is used to input the acquired multi-index continuous sampling data sequence of a certain monitoring station into the coding network structure of the variational autoencoder to obtain the first reconstruction probability of the mean and standard deviation distribution of each input index data. The network module is also used to input normal indicator data into the coding network structure of the variational autoencoder to obtain the second reconstruction probability of the distribution of the corresponding indicators of the normal data. The data comparator is used to compare the magnitude of the first reconstruction probability and the second reconstruction probability. If the first reconstruction probability is less than the second reconstruction probability, and the product of the first reconstruction probability and the second reconstruction probability is less than a preset probability threshold, then it is considered that multiple indicators of a certain monitoring station have become abnormal; otherwise, it is considered that multiple indicators of a certain monitoring station have not become abnormal.

5. The diagnostic system for simultaneous anomalies of multiple indicators based on reconstruction probability according to claim 4, characterized in that, The system also includes: A data processor is used to calculate a comprehensive anomaly score for multiple indicators if multiple indicators at a monitoring station show abnormalities.

6. The diagnostic system for simultaneous anomalies of multiple indicators based on reconstruction probability according to claim 4, characterized in that, The network module is configured with a variational autoencoder encoding network structure; The coding network structure of the variational autoencoder includes an encoder and a decoder.

Images