A collection system risk assessment method, system, medium, equipment and program

Abstract

The present invention relates to the technical field of electricity information collection, and provides a collection system risk assessment method, system, medium, device and program. The collection system risk assessment method includes: using a game theory combined weighting method based on the historical operation data of the collection system to obtain a risk assessment index combination vector; based on the demand information of the power business expansion, using a convolutional neural network to perform representation learning of the power business expansion demand; using a Transformer mechanism to capture the correlation between the risk assessment index and the power business expansion demand information, and using an autoencoder to obtain the uncertainty information brought by the power business expansion; using cross attention to obtain a comprehensive representation vector of the collection system operation status; by using a strategy combining a bidirectional gated recurrent unit with contrastive learning, the risk assessment of the collection system operation status is realized, and the accuracy of the collection system risk assessment is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electricity consumption information collection, and in particular to a collection system risk assessment method, system, medium, equipment and program. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] The expansion of power business covers real-time monitoring and analysis of user needs, market changes and technological innovations. Based on the widespread application of distributed generation, demand-side management and renewable energy, users' electricity consumption patterns and demand changes are becoming more complex. This complexity requires power companies to have more agile and efficient information response capabilities to meet users' personalized electricity needs. As the front end of information acquisition, the performance of the power information collection system directly affects the decision-making ability and response speed of power companies. Traditional power information collection systems are mostly focused on the metering of electricity, and their functions are relatively single, which cannot meet the requirements of the increasingly complex power market environment. In addition, traditional power information collection systems are often unable to adapt to these changes, resulting in the accuracy and real-time performance of data collection being affected. Summary of the invention

[0004] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a collection system risk assessment method, system, medium, equipment and program. The present invention integrates the historical operation data of the electricity consumption information collection system and the demand information of the expansion of electricity business, captures the uncertainty information brought by the expansion of electricity business, and deeply mines the potential information behind the electricity data. It can provide strong support for power management, load forecasting and electricity price setting, and provide more effective services for power companies.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] A first aspect of the present invention provides a collection system risk assessment method.

[0007] A collection system risk assessment method, comprising:

[0008] The historical operation data of the electricity consumption information collection system is obtained, and the game weighted combination method is used to obtain the risk assessment indicator combination vector; among which, the historical operation data of the electricity consumption information collection system includes current fluctuations, voltage stability, collection frequency, response time, equipment status and other data in different time periods.

[0009] Obtain demand information for power business expansion, and use convolutional neural networks to obtain a representation vector for power business expansion demand;

[0010] Based on the risk assessment index combination vector and the power business expansion demand representation vector, the Transformer mechanism is used to learn the correlation between the two vectors to obtain the implicit representation vector of the system operation status.

[0011] Based on the demand information of power business expansion, a variational autoencoder is used to obtain the uncertainty representation vector caused by power business expansion.

[0012] Based on the implicit representation vector of the system operating state and the uncertainty representation vector caused by the expansion of power business, a cross-attention mechanism is used to quantify the impact of uncertainty information on the system operating state, and a comprehensive representation vector of the system operating state is obtained.

[0013] Based on the comprehensive representation vector of the system operation status, a bidirectional gated recurrent unit neural network is used to obtain the risk assessment vector of the system operation status;

[0014] Based on the risk assessment vector of the system operating status, the loss function is calculated, and the learning parameters of the model are trained using the back propagation algorithm; the trained model is used to evaluate the risk of the operating status of the electricity consumption information collection system.

[0015] Furthermore, after the model training is completed, the test sample set is used to perform predictive testing, and the test results are compared with the operating thresholds set by the actual power collection system. The underlying data information is updated based on the feedback of the comparison results, so as to continuously optimize the model weight value and continuously improve the risk assessment of the collection system operation status.

[0016] Furthermore, the historical operation data of the power consumption information collection system is obtained, and a game weighted combination method is used to obtain a risk assessment indicator combination vector; the method includes:

[0017] The historical operation data of the electricity consumption information collection system is standardized to obtain a standardized data matrix to calculate the covariance matrix; the eigenvalue problem of the covariance matrix is ​​solved to obtain the eigenvalue and the corresponding eigenvector ; According to the characteristic value The size of the feature vector Sort and select the first K eigenvectors as principal components to obtain the objective weight vector F of the evaluation index;

[0018] Calculate the entropy value of each indicator factor in the electricity consumption information collection system to calculate the weight of each indicator and obtain the subjective weight vector G of the evaluation indicator;

[0019] Based on the objective weight vector F of the evaluation index and the subjective weight vector G of the evaluation index, the game weighting method is used to analyze the risk assessment indicators that affect the operating status of the electricity consumption information collection system, and the risk assessment indicator combination vector is obtained.

[0020] Furthermore, for the objective weight vector F of the evaluation indicators and the subjective weight vector G of the evaluation indicators, the game weighting method is used to analyze the risk assessment indicators affecting the operation status of the power consumption information collection system, and a risk assessment indicator combination vector is obtained. The method includes:

[0021] Perform a consistency test on the objective weight vector F of the evaluation indicators and the subjective weight vector G of the evaluation indicators;

[0022] Through the consistency test, construct a basic weight set;

[0023] Using the idea of game theory, find the optimal weight vector among all possible basic weight sets to analyze the risk assessment indicators affecting the operation status of the power consumption information collection system, and obtain a risk assessment indicator combination vector.

[0024] Furthermore, for the demand information based on the expansion of power services, a variational autoencoder is used to obtain an uncertainty representation vector brought about by the expansion of power services. The method includes:

[0025] Use an encoder to map the demand information of the expansion of power services to latent variables;

[0026] Use a decoder to map the latent variables back to the data space and output an uncertainty representation vector brought about by the expansion of power services. Among them, the variational autoencoder includes an encoder and a decoder.

[0027] Furthermore, the loss function is represented by the following formula:

[0028]

[0029] Among them, represents the loss function, represents the contrastive learning loss function, represents the cross-entropy loss function, and are hyperparameters, represents the set of learnable parameters in the risk assessment model.

[0030] Furthermore, after obtaining the historical operation data of the power consumption information collection system, preprocess the data. The preprocessing process includes missing value marking, non-missing value standardization, and define and store the data.

[0031] The second aspect of the present invention provides a risk assessment system for a collection system.

[0032] A risk assessment system for a collection system includes:

[0033] The system operation risk assessment index combination vector acquisition module is used to obtain the historical operation data of the power consumption information collection system and obtain the risk assessment index combination vector by using the game weighted combination method;

[0034] The power business expansion demand representation vector acquisition module is used to obtain the demand information of the power business expansion and obtain the power business expansion demand representation vector by using a convolutional neural network;

[0035] The implicit vector acquisition module of the system operation status is used to learn the correlation between the risk assessment index combination vector and the power business expansion demand representation vector using the Transformer mechanism to obtain the implicit representation vector of the system operation status;

[0036] The power business expansion uncertainty representation vector acquisition module is used to obtain the uncertainty representation vector based on the demand information of the power business expansion by using a variational autoencoder;

[0037] A system operation state comprehensive representation vector acquisition module is used to quantify the impact of uncertainty information on the system operation state based on the implicit representation vector of the system operation state and the uncertainty representation vector, and obtain a comprehensive representation vector of the system operation state by using a cross-attention mechanism;

[0038] A system operation status risk assessment vector acquisition module is used to obtain a risk assessment vector of the system operation status based on a comprehensive representation vector of the system operation status by using a bidirectional gated recurrent unit neural network;

[0039] The system operation status risk assessment module is used to calculate the loss function based on the risk assessment vector of the system operation status, and use the back propagation algorithm to train the learning parameters of the model; and use the trained model to evaluate the risk of the operation status of the power consumption information collection system.

[0040] A third aspect of the present invention provides a computer-readable storage medium.

[0041] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps in the acquisition system risk assessment method as described in the first aspect above.

[0042] A fourth aspect of the present invention provides a computer device.

[0043] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the acquisition system risk assessment method as described in the first aspect above are implemented.

[0044] A fifth aspect of the present invention provides a computer program product or a computer program.

[0045] The present invention provides a computer program product or a computer program, the computer program product or the computer program comprising computer instructions, the computer instructions being stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the steps in the acquisition system risk assessment method as described in the first aspect above.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] (1) Based on the historical operation data of the acquisition system and the expansion demand information of power business, the present invention adopts game empowerment, convolutional neural network, Transformer and other methods to integrate the correlation between the two information, deeply explore the potential value behind the power data, improve the requirements of different businesses for various data resources and performance of the acquisition system, and effectively coordinate the allocation of power resources.

[0048] (2) Based on the demand information of power business expansion, the present invention adopts variational autoencoder and cross-attention mechanism to quantify the impact of uncertainty information on the system operation status, and realizes the risk assessment of the system operation status by adopting a strategy combining bidirectional gated recurrent unit neural network and contrastive learning. This not only enhances the modeling ability of complex power system business relationships, but also improves the adaptability of the system in the face of changing environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0050] Figure 1 is a flow chart of a collection system risk assessment method provided by an embodiment of the present invention;

[0051] Figure 2 is a flow chart of a system operation status risk assessment embodiment provided by an embodiment;

[0052] Figure 3 is a graph showing changes in the training loss value of the acquisition system risk assessment method provided in the embodiment;

[0053] Figure 4 It is a structural diagram of the acquisition system risk assessment system provided in the embodiment. DETAILED DESCRIPTION

[0054] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0055] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0056] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0057] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the methods and systems according to various embodiments of the present disclosure. It should be noted that each box in the flowchart or block diagram can represent a module, a program segment, or a part of a code, and the module, program segment, or a part of a code may include one or more executable instructions for implementing the logical functions specified in each embodiment. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, or they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the flowchart and / or block diagram, and the combination of boxes in the flowchart and / or block diagram can be implemented using a dedicated hardware-based system that performs a specified function or operation, or can be implemented using a combination of dedicated hardware and computer instructions.

[0058] Embodiment 1

[0059] like Figure 1 , Figure 2 As shown, this embodiment provides a collection system risk assessment method, comprising the following steps:

[0060] Step 1: Collect the status data of each link of the electricity consumption information collection system and preprocess the data, including missing value marking, non-missing value standardization, data definition and storage, etc.

[0061] Specifically, the operating status data of the power consumption information collection system of a provincial company of State Grid is used as the data set. The data set includes the operating data of each link from June 25, 2023 to July 15, 2024, which is collected every hour. Among them, the operating data of each link of the power consumption information collection system includes current fluctuations, voltage stability, collection frequency, response time, equipment status and other data in different time periods. After data preprocessing and dimensionality reduction, the data set is divided into training set, validation set and test set in a ratio of 7:1.5:1.5.

[0062] Step 2: In this embodiment, the time series data of the historical operation of the electricity consumption information collection system is defined as , represents the time series data of T moments. For a certain time t, , where M represents the number of operating links of the current electricity consumption information collection system. Represents the time t The characteristic attributes of meter reading data at each link (such as current, voltage, throughput, response time, etc.)

[0063] Step 3: Based on the historical operation data of the electricity consumption information collection system, the principal component analysis method and the entropy weight method are first used to obtain the objective weight vector and the subjective weight vector of the evaluation index respectively. Then, the game empowerment weight combination method is used to analyze the risk assessment indicators that affect the operation status of the collection system, and the risk assessment indicator combination vector is obtained, which mainly includes data collection rate, data processing delay, data accuracy, system stability, user load capacity and data processing capacity, response time, etc. The relevant situation is shown in Table 1.

[0064] Table 1 Risk assessment indicators and their weights affecting the operation status of power systems

[0065]

[0066] Specifically, the general generation process of obtaining the risk assessment index combination vector based on the game combination weighting method in step 3 is:

[0067] Step 3.1: First, based on the business objects and object attributes in the entire operation process of the electricity consumption information collection system, the historical operation data is standardized to eliminate the differences in dimensions and magnitudes between variables and make each variable comparable. The calculation formula is as follows:

[0068]

[0069] in, is the standardized data, X is the historical operation data of the acquisition system, is the mean, is the standard deviation.

[0070] Then, based on the standardized data matrix , calculate the covariance matrix , reflecting the degree of correlation between variables. Covariance matrix The calculation formula is as follows:

[0071]

[0072] in, represents the number of samples, the covariance matrix Describes the correlation between feature variables.

[0073] Next, we use matrix knowledge to solve the covariance matrix The eigenvalue problem of and the corresponding eigenvector .

[0074] Finally, the eigenvectors are sorted according to the size of the eigenvalues, and the first K eigenvectors are selected as principal components to obtain the objective weight vector F of the evaluation index.

[0075] Step 3.2: Based on the historical operation data of the collection system, the entropy weight method is used to obtain the subjective weight vector of the evaluation index. The entropy weight method is a weight calculation method based on information entropy, which mainly determines the relative importance of each indicator in the comprehensive evaluation by calculating its entropy value.

[0076] First, calculate the entropy value of each indicator factor in the operation status of the acquisition system :

[0077]

[0078] in, is the weight of the jth sample under the i-th index, that is, , n is the number of samples, It is the specific value of the j-th sample under the i-th indicator.

[0079] Then, the entropy values ​​of each indicator factor are normalized:

[0080]

[0081] in, It represents the entropy value of each indicator factor after normalization.

[0082] Finally, calculate the weight of each indicator , thereby obtaining the subjective weight vector of the evaluation index :

[0083]

[0084] in, is the number of indicators.

[0085] Step 3.3: Based on the objective weight vector F and subjective weight vector G of the evaluation indicators obtained in steps 3.1 and 3.2, the game weighting method is used to analyze the risk assessment indicators that affect the operating status of the acquisition system and obtain the risk assessment indicator combination vector.

[0086] In the game theory combined weighting method, each evaluation indicator is assigned a different weight to reflect its importance and influence in the risk assessment process. Through the combination of different weights and game analysis, the final decision result can be obtained so that each evaluation indicator can obtain the maximum benefit. The steps of weight combination based on game weighting are as follows:

[0087] 1) Consistency check

[0088] Before combining weights, the weights obtained by each weighting method should be checked for consistency to prevent the results obtained by different methods from being too contradictory. ,like:

[0089]

[0090] in, Indicates Indicators, when for any indicator When all of them are satisfied, it means that the obtained weights pass the consistency test. Indicates The weight vector of the method.

[0091] 2) Through consistency check, based on the objective weight vector F and the subjective weight vector G, let the coefficients of the objective weight vector and the subjective weight vector be and , then the linear combination of the two vector weights It is expressed as:

[0092]

[0093] 3) Using the idea of ​​game theory, the goal is to minimize the distance difference. and Optimize, that is:

[0094]

[0095] in, .

[0096] 4) Finally, based on the above game weighting idea, the optimized weight coefficients are obtained, which are and , thereby calculating the risk assessment index combination vector The calculation formula is as follows:

[0097]

[0098] Step 4: Based on the demand information of power business expansion, a method combining convolutional neural network and residual network is used to learn the representation of power business expansion demand and obtain the power business expansion demand representation vector.

[0099] Convolution calculation is an important operation in convolutional neural networks. It performs representation vector learning by setting convolution kernels on the extended demand information of power business. The convolution calculation formula is as follows:

[0100]

[0101] in, It is the demand information for the expansion of power business; The representation vector is obtained through convolution operation; is the convolution kernel; c is the size of the convolution kernel; the subscript in the formula and Indicates the location of the output, and Indicates the location of the convolution kernel.

[0102] Based on the representation vector obtained by the convolutional neural network, this embodiment introduces a residual network to improve the network's ability to capture information, thereby better learning the representation vector of the business expansion requirements The specific formula is as follows:

[0103]

[0104] Step 5: Risk assessment indicator combination vector obtained based on step 3 And the power business expansion demand representation vector obtained in step 4 , the Transformer mechanism is used to learn the relationship between the two vectors and obtain the implicit representation vector of the system operation status. The specific formula is as follows:

[0105] ,

[0106]

[0107]

[0108]

[0109] in, , , V represent the query vector, key vector and value vector in the Transformer structure respectively; , , , They all represent weight parameter matrices, represents the head vector, represents the embedding dimension, represents the deviation matrix; Concat () indicates the concatenation operation; Norm () represents the normalization function; FFN () represents a feedforward neural network; Represents the implicit representation vector of the system operation status obtained through Transformer.

[0110] Step 6: Based on the demand information of power business expansion, a variational autoencoder is used to capture the uncertainty information brought by the expansion of power business and obtain the uncertainty representation vector.

[0111] Variational autoencoders are a type of generative model that can learn potential attributes and construct new elements from the probability distribution of latent variable space to obtain uncertainty information. Variational autoencoders mainly involve two parts: encoder and decoder.

[0112] The purpose of the encoder is to convert power business requirements Mapped to the latent variable Z, the implementation process is as follows:

[0113]

[0114] in, and represents the mean and variance, is from the labeled normal distribution The noise sampled in .

[0115] The purpose of the decoder is to map the latent variables Z back to the data space and output an uncertainty vector U:

[0116]

[0117] in, represents the mapping function, represents random parameters.

[0118] Step 7: Based on the implicit representation vector of the system operation status obtained in step 5 The business expansion uncertainty representation vector U obtained in step 6 uses the cross attention mechanism to quantify the impact of uncertainty information on the system operation state, thereby obtaining a comprehensive representation vector of the system operation state ;

[0119]

[0120]

[0121] in, represents the query vector; represents the parameter matrix, Represents the attention score.

[0122] Step 8: Based on the comprehensive representation vector of the system operation status obtained in step 7, a bidirectional gated recurrent unit neural network (BiGRU) is used to output the risk assessment vector of the system operation status. The implementation of the standard GRU is as follows:

[0123]

[0124]

[0125]

[0126]

[0127] in, , , and They are the hidden state, reset gate, update gate and intermediate state of GRU output respectively. , , and , , are learnable parameters. It means multiplying the corresponding elements in these two matrices.

[0128] Based on the calculation process of the above standard GRU, this embodiment uses the result obtained by the bidirectional GRU calculation as the risk assessment vector of the system operation status. The implementation process is as follows:

[0129]

[0130]

[0131]

[0132] in, is the hidden state vector obtained by the forward GRU; Represents forward GRU, that is, calculation from front to back; is the hidden state vector obtained by the reverse GRU; Represents reverse GRU, that is, calculation from back to front; is the comprehensive representation vector of the system operation status obtained in step 7. Finally, the hidden state vectors obtained in two directions are concatenated to obtain the final risk assessment vector of the system operation status .

[0133] Step 9: Based on the system operation status risk assessment vector obtained in step 8 ,This implementation adopts a contrastive learning strategy to further enhance the representation ability of the risk assessment vector. The core idea of ​​contrastive learning is to learn feature representation by comparing the similarities between samples. It aims to distinguish similar samples from dissimilar samples by grouping them together. Specifically, the loss function of contrastive learning is It can be expressed as:

[0134]

[0135] in, is a function that measures the correlation between two representation vectors with cosine similarity, represents the number of positive and negative sample pairs, is the temperature parameter in the softmax function, Represents the risk assessment representation vector learned from negative samples.

[0136] Step 10: The risk assessment vector of the system operation status obtained in step 8 , input the softmax function to conduct risk assessment of the operation status of the electricity consumption information collection system:

[0137]

[0138] in, represents the system operation status risk assessment value output by the model, and is the learning parameter of the softmax function.

[0139] Step 11: Use cross entropy as the loss function for model discrimination. If y is the true category distribution, the loss function is defined as follows:

[0140]

[0141] in, represents the total number of samples, Indicates l The risk assessment value of a sample, Indicates l The true value of the samples, Represents the calculated cross entropy loss value.

[0142] Step 12: Through steps 9 and 6, the overall loss function of the model is defined as follows:

[0143]

[0144] in, is the overall loss function value of the model. and is a hyperparameter. Represents the set of learnable parameters in the risk assessment model. It represents L2 regularization, which can effectively prevent the model from overfitting. Figure 3 The figure shows the change of loss value during the model training process of this embodiment. Through the iterative solution of step 9 and step 6, the loss value is continuously minimized. , thereby obtaining the learning parameters of the model proposed in this embodiment.

[0145] The test sample is subjected to a risk assessment of the operation status of the power consumption information collection system, and the assessment results are pushed and compared with the threshold set by the actual power consumption information collection system. In order to verify the performance of the model, this embodiment uses accuracy, precision, recall rate and F1 score as evaluation indicators of risk assessment. The larger the values ​​of these four evaluation indicators, the better the performance of the model. The specific calculation formula is as follows:

[0146]

[0147]

[0148]

[0149]

[0150] Among them, Total is the total number of samples, TP (True Positive) is the number of abnormal samples correctly evaluated, FP (False Positive) is the number of abnormal samples incorrectly evaluated, TN (True Negative) is the number of normal samples correctly evaluated, and FN (False Negative) is the number of normal samples incorrectly evaluated.

[0151] The experimental parameter settings of this embodiment are shown in Table 2.

[0152] Table 2 Parameter settings

[0153]

[0154] The performance comparison of the risk assessment method in this embodiment is shown in Table 3.

[0155] Table 3 Performance comparison of different risk assessment methods

[0156]

[0157] Based on the results in Table 3, it can be seen that the performance of the acquisition system operation status risk assessment proposed in this embodiment is better than other methods.

[0158] Embodiment 2

[0159] like Figure 4 As shown, this embodiment provides a collection system risk assessment system, including:

[0160] A system operation risk assessment index combination vector acquisition module is used to analyze the risk assessment indicators that affect the operation status of the collection system based on the historical operation data of the power consumption information collection system by using a game-based weighted combination method to obtain a risk assessment index combination vector;

[0161] A power business expansion demand representation vector acquisition module is used to perform representation learning of power business expansion demand based on demand information of power business expansion by using a convolutional neural network to obtain a power business expansion demand representation vector;

[0162] The implicit vector acquisition module of the system operation status is used to learn the correlation between the risk assessment index vector and the power business expansion demand representation vector based on the obtained risk assessment index vector and the power business expansion demand representation vector, and obtain the implicit representation vector of the system operation status by using the Transformer mechanism;

[0163] The power business expansion uncertainty representation vector acquisition module is used to capture the uncertainty information brought by the power business expansion based on the demand information of the power business expansion, and obtain the uncertainty representation vector by using the variational autoencoder;

[0164] A system operation status comprehensive representation vector acquisition module is used to quantify the impact of uncertainty information on the system operation status based on the obtained implicit representation vector of the system operation status and the service extension uncertainty representation vector, thereby obtaining a comprehensive representation vector of the system operation status;

[0165] A system operation status risk assessment vector acquisition module, which is used to output the risk assessment vector of the system operation status by using a bidirectional gated recurrent unit neural network based on the obtained system operation status comprehensive representation vector;

[0166] The system operation status risk assessment module is used to utilize the risk assessment vector of the system operation status, adopt the contrastive learning method, construct the softmax function, calculate the loss function of the output value of the softmax function, and use the back propagation algorithm to train the learning parameters of the model to complete the risk assessment of the operation status of the power consumption information collection system.

[0167] The feedback update module is used to output the evaluation results of the test sample set after the model training is completed, compare them with the thresholds set by the actual power consumption collection system, and feedback to update the underlying data information, so as to continuously optimize the model weight value and continuously improve the risk assessment of the system operation status.

[0168] In some embodiments, the system further includes: a preprocessing module for performing data preprocessing on the historical operation data of the electricity consumption information collection system, including standardization of non-missing value features, labeling of missing value features, etc.

[0169] Embodiment 3

[0170] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps in the acquisition system risk assessment method described in the first embodiment are implemented.

[0171] Embodiment 4

[0172] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps in the acquisition system risk assessment method described in the first embodiment are implemented.

[0173] Embodiment 5

[0174] This embodiment provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the steps in the acquisition system risk assessment method described in the first embodiment.

[0175] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.

[0176] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0177] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0178] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0179] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0180] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A collection system risk assessment method, characterized in that: include: Obtain the historical operation data of the electricity consumption information collection system, and use the game weighted combination method to obtain the risk assessment indicator combination vector; Obtain demand information for power business expansion, and use convolutional neural networks to obtain a representation vector for power business expansion demand; Based on the risk assessment index combination vector and the power business expansion demand representation vector, the Transformer mechanism is used to learn the correlation between the two vectors to obtain the implicit representation vector of the system operation status. Based on the demand information of power business expansion, a variational autoencoder is used to obtain the uncertainty representation vector caused by power business expansion. The method of obtaining the uncertainty representation vector caused by the expansion of electric power services by using a variational autoencoder based on the demand information of the expansion of electric power services includes: An encoder is used to map the demand information of power business expansion to latent variables; A decoder is used to map the latent variables back to the data space, and an uncertainty representation vector caused by the expansion of the power business is output; wherein the variational autoencoder includes an encoder and a decoder; Based on the implicit representation vector of the system operating state and the uncertainty representation vector caused by the expansion of power business, a cross-attention mechanism is used to quantify the impact of uncertainty information on the system operating state, and a comprehensive representation vector of the system operating state is obtained. Based on the comprehensive representation vector of the system operation status, a bidirectional gated recurrent unit neural network is used to obtain the risk assessment vector of the system operation status; Based on the risk assessment vector of the system operating status, the loss function is calculated, and the learning parameters of the model are trained using the back propagation algorithm; the trained model is used to evaluate the risk of the operating status of the electricity consumption information collection system.

2. The acquisition system risk assessment method according to claim 1, characterized in that: After the model training is completed, the test sample set is used to perform predictive testing, and the test results are compared with the operating thresholds set by the actual electricity consumption collection system. The underlying data information is updated based on the comparison results and the model weight values ​​are optimized.

3. The acquisition system risk assessment method according to claim 1 or 2, characterized in that: The historical operation data of the power consumption information collection system is obtained, and a risk assessment index combination vector is obtained by adopting a game weighted combination method; the method includes: The historical operation data of the electricity consumption information collection system is standardized to obtain a standardized data matrix to calculate the covariance matrix; the eigenvalue problem of the covariance matrix is ​​solved to obtain the eigenvalue and the corresponding eigenvector ; According to the characteristic value The size of the feature vector Sort and select the first K eigenvectors as principal components to obtain the objective weight vector F of the evaluation index; Calculate the entropy value of each indicator factor in the electricity consumption information collection system to calculate the weight of each indicator and obtain the subjective weight vector G of the evaluation indicator; Based on the objective weight vector F of the evaluation index and the subjective weight vector G of the evaluation index, the game weighting method is used to analyze the risk assessment indicators that affect the operating status of the electricity consumption information collection system, and the risk assessment indicator combination vector is obtained.

4. The acquisition system risk assessment method according to claim 1 or 2, characterized in that: The loss function is expressed by the following formula: in, represents the loss function, represents the contrastive learning loss function, represents the risk assessment loss function, and is a hyperparameter, Represents the set of learnable parameters in the risk assessment model.

5. The acquisition system risk assessment method according to claim 1 or 2, characterized in that: After the historical operation data of the electricity consumption information collection system is obtained, the data is preprocessed, and the preprocessing process includes missing value marking, non-missing value standardization, and the data is defined and stored.

6. A collection system risk assessment system, characterized in that: include: The system operation risk assessment index combination vector acquisition module is used to obtain the historical operation data of the power consumption information collection system and obtain the risk assessment index combination vector by using the game weighted combination method; The power business expansion demand representation vector acquisition module is used to obtain the demand information of the power business expansion and obtain the power business expansion demand representation vector by using a convolutional neural network; The implicit vector acquisition module of the system operation status is used to learn the correlation between the risk assessment index combination vector and the power business expansion demand representation vector using the Transformer mechanism to obtain the implicit representation vector of the system operation status; The power business expansion uncertainty representation vector acquisition module is used to obtain the uncertainty representation vector caused by the power business expansion based on the demand information of the power business expansion by using the variational autoencoder; The method of obtaining the uncertainty representation vector caused by the expansion of electric power services by using a variational autoencoder based on the demand information of the expansion of electric power services includes: An encoder is used to map the demand information of power business expansion to latent variables; A decoder is used to map the latent variables back to the data space, and an uncertainty representation vector caused by the expansion of the power business is output; wherein the variational autoencoder includes an encoder and a decoder; The system operation state comprehensive representation vector acquisition module is used to quantify the impact of uncertainty information on the system operation state based on the implicit representation vector of the system operation state and the uncertainty representation vector caused by the expansion of power business, and obtain the comprehensive representation vector of the system operation state by using the cross attention mechanism; A system operation status risk assessment vector acquisition module is used to obtain a risk assessment vector of the system operation status based on a comprehensive representation vector of the system operation status by using a bidirectional gated recurrent unit neural network; The system operation status risk assessment module is used to calculate the loss function based on the risk assessment vector of the system operation status, and use the back propagation algorithm to train the learning parameters of the model; and use the trained model to evaluate the risk of the operation status of the power consumption information collection system.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in the acquisition system risk assessment method as described in any one of claims 1 to 5 are implemented.

8. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the acquisition system risk assessment method according to any one of claims 1 to 5 are implemented.

9. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps in the acquisition system risk assessment method according to any one of claims 1 to 5 are implemented.

Images