CN-122026622-A - Electric power intelligent fusion terminal

CN122026622ACN 122026622 ACN122026622 ACN 122026622ACN-122026622-A

Abstract

The invention discloses an electric power intelligent fusion terminal, which relates to the technical field of intelligent power grids and comprises a multi-mode acquisition module, an execution optimization module and an intelligent fusion terminal, wherein the multi-mode acquisition module acquires electric quantity data and temperature monitoring data, integrates and generates a multi-mode operation data set, performs data cleaning on the multi-mode operation data set, calculates zero sequence voltage abrupt change, three-phase current unbalance and temperature change gradient characteristic quantity and generates a standardized characteristic vector, the intelligent fusion terminal executes control operation corresponding to a collaborative consensus instruction to generate an execution result, packages the execution result, an interpretable knowledge packet and all signature feedback information into event logs, uploads the event logs to a cloud master station, and issues causal inference model optimization parameters through the cloud master station to dynamically update a causal inference model to complete iterative optimization. The invention improves the reliability and safety of decision making through a multi-terminal collaborative verification mechanism.

Inventors

CHEN JIE
YU XINWEI
XU LINGXIANG
YOU XIAOBO
GONG YU
Wu Qiuya
Jin Xiaofu

Assignees

杭州百富电子技术有限公司

Dates

Publication Date: 20260512
Application Date: 20260104

Claims (10)

1. An intelligent electric power fusion terminal is characterized by comprising, The multi-mode acquisition module is used for acquiring electric quantity data and temperature monitoring data, integrating and generating a multi-mode operation data set, carrying out data cleaning on the multi-mode operation data set, calculating zero sequence voltage abrupt change, three-phase current unbalance and temperature change gradient characteristic quantity, and generating a standardized characteristic vector; The intelligent research and judgment module inputs the standardized feature vector into a causal inference model, deduces the root cause and potential result of the abnormal state of the power grid, generates a decision result and causal evidence chain, compresses the decision result and causal evidence chain into an interpretable knowledge packet, extracts causal feature data, and calculates a causal feature data hash value; The collaborative verification module broadcasts the interpretable knowledge packet and the causal feature data hash value to the adjacent intelligent fusion terminal through the wireless communication interface, triggers a collaborative verification timer, verifies the integrity of causal feature data, calls historical causal feature data, and generates signature feedback information; the consensus achieving module is used for verifying the passing rate of all signature feedback information before the cooperative verification timer is overtime, and generating a cooperative consensus instruction when the passing rate exceeds a preset passing threshold; And the execution optimization module is used for executing control operation corresponding to the collaborative consensus instruction by the intelligent fusion terminal, generating an execution result, packaging the execution result, the interpretable knowledge packet and all signature feedback information into an event log, uploading the event log to the cloud master station, issuing causal inference model optimization parameters by the cloud master station, dynamically updating the causal inference model, and completing iterative optimization.
2. The intelligent power fusion terminal of claim 1, wherein the electrical quantity data comprises three-phase voltage data, three-phase current data and zero sequence voltage data; The temperature monitoring data comprise distribution transformer oil temperature data and terminal box body internal temperature data.
3. The intelligent power fusion terminal of claim 1, wherein the generating the multi-modal operational data set comprises the steps of, Performing feature extraction on three-phase voltage data, three-phase current data and zero sequence voltage data in the electric quantity data to generate an electric quantity feature vector; extracting characteristics of the temperature data of the distribution transformer oil in the temperature monitoring data and the temperature data in the terminal box body to generate a temperature characteristic vector; The sampling frequency of the electrical quantity characteristic vector is kept consistent with that of the temperature characteristic vector through interpolation processing, so that time synchronization is realized; and carrying out time alignment combination on the synchronized electrical quantity characteristic vector and the synchronized temperature characteristic vector to generate a multi-mode operation data set.
4. The intelligent power fusion terminal according to claim 1, wherein the generating of the standardized feature vector comprises the following steps, Performing outlier detection and cleaning treatment on the multi-mode operation data set by adopting a sliding window quadripotential method to generate a cleaned multi-mode operation data set; based on the cleaned multi-mode operation data set, calculating a zero sequence voltage abrupt change, three-phase current unbalance degree and temperature change gradient characteristic quantity, and carrying out characteristic level fusion to form an original characteristic vector; And calling the historical quantile reference value to perform standardized conversion on the original feature vector, and generating a standardized feature vector.
5. The intelligent power fusion terminal according to claim 1, wherein the decision result and causal evidence chain is generated by the following steps, Inputting the standardized feature vector into a causal inference model, generating a causal intensity matrix among variables through time sequence causal discovery analysis, and constructing a causal graph structure with weights based on the causal intensity matrix; executing inverse fact reasoning analysis on the causal graph structure, collecting an actual observed value, and identifying root causes and potential consequences by comparing the actual observed value with the abnormal state difference of the power grid under the inverse fact condition to generate a decision result; And analyzing the causal relation and the influence degree of the causal graph structure and the decision result to generate a causal evidence chain.
6. The intelligent power fusion terminal according to claim 1, wherein the calculating of the causal characteristic data hash value comprises the following steps, Constructing a decision-evidence associated hypergraph structure according to the decision result and a causal evidence chain, and performing neural network aggregation processing on the hypergraph structure to generate graph node semantic codes; compressing the semantic codes of the graph nodes through knowledge distillation to generate a light interpretable knowledge packet, and extracting causal feature data from the interpretable knowledge packet; And calculating the hash value of the causal characteristic data by adopting a secure hash algorithm on the causal characteristic data.
7. The intelligent power fusion terminal according to claim 1, wherein the triggering of the cooperative authentication timer comprises the following steps, Acquiring network topology information and channel state information through signaling interaction between terminals, and determining wireless broadcast parameters by utilizing dynamic spectrum access optimization; Encoding the interpretable knowledge packet and the causal characteristic data hash value into a multi-user composite signal with the same frequency as the wireless broadcast parameter by adopting a non-orthogonal multiple access mechanism, and broadcasting the multi-user composite signal to an adjacent intelligent fusion terminal through a wireless communication interface; and dynamically calculating the timeout time of the collaborative authentication timer according to the network condition and the task emergency degree and triggering the collaborative authentication timer when broadcasting is started.
8. The intelligent power fusion terminal according to claim 1, wherein the signature feedback information is generated by the steps of, Receiving interpretable knowledge packets and causal characteristic data hash values from adjacent intelligent fusion terminals, and verifying the integrity of the causal characteristic data through the same secure hash algorithm; according to the causal feature data passing verification, historical causal feature data in a historical verification knowledge packet is called to carry out semantic consistency analysis, and a semantic consistency analysis result is generated; and carrying out credibility weighting and time attenuation processing on the semantic consistency analysis result to generate signature feedback information.
9. The intelligent power fusion terminal according to claim 1, wherein the generating of the collaborative consensus instruction comprises the following steps, Before the cooperative verification timer is overtime, collecting signature feedback information returned by all adjacent intelligent fusion terminals; carrying out digital signature verification on all signature feedback information, and screening out signature feedback information passing verification; And calculating the ratio of the number of signature feedback information passing verification to the number of total signature feedback information, acquiring the passing rate, and generating a cooperative consensus instruction when the passing rate exceeds a preset passing threshold.
10. The intelligent power fusion terminal of claim 1, wherein the dynamically updated causal inference model performs iterative optimization by the steps of, Executing control operation corresponding to the collaborative consensus instruction, generating an execution result, constructing an execution result, an interpretable knowledge packet and all signature feedback information into an event log which cannot be tampered, and uploading the event log to a cloud master station; the cloud master station generates causal inference model optimization parameters according to the non-tamperable event log and transmits the causal inference model optimization parameters to the intelligent fusion terminal through a privacy protection mechanism; And dynamically updating network weights and parameters of the causal inference model according to the issued causal inference model optimization parameters, and completing an iterative optimization loop through performance evaluation.

Description

Electric power intelligent fusion terminal Technical Field The invention relates to the technical field of smart power grids, in particular to an intelligent power fusion terminal. Background With the rapid development of smart grids, power distribution network state monitoring has evolved gradually from single electric quantity acquisition to multi-mode data fusion analysis. In recent years, the application of the sensing technology of the internet of things, the edge computing architecture and the artificial intelligence algorithm in the electric power intelligent fusion is mature, so that the terminal equipment has the capability of synchronously acquiring and primarily processing multidimensional data such as electric quantity, temperature, environmental parameters and the like. The prior art can realize fault early warning based on a rule mechanism or a machine learning model, and complete centralized storage and offline analysis of data through a cloud edge cooperative mechanism. However, the prior art has the defects in the aspects of multi-terminal collaborative decision and real-time performance, the traditional method relies on cloud centralized calculation, so that decision delay is larger, the requirements of rapid isolation and recovery of power distribution network faults are difficult to meet, and the lack of a safe and reliable collaborative verification mechanism between terminals can not ensure the reliability and consistency of distributed decisions, so that misjudgment or response conflict is easy to occur particularly in a high-concurrency fault scene. Disclosure of Invention The present invention has been made in view of the above-described problems occurring in the prior art. Therefore, the invention provides the intelligent electric power fusion terminal for solving the problems of real-time performance, reliability and safety of multi-terminal collaborative decision under the abnormal state of the power distribution network. In order to solve the technical problems, the invention provides the following technical scheme: The invention provides an electric power intelligent fusion terminal which comprises a multi-mode acquisition module, a cooperative verification module, an intelligent judging module, an execution optimization module, a master station, a communication module and a communication module, wherein the multi-mode acquisition module acquires electric quantity data and temperature monitoring data, integrates and generates a multi-mode operation data set, performs data cleaning on the multi-mode operation data set, calculates zero sequence voltage abrupt change, three-phase current unbalance and temperature change gradient characteristic quantity, generates a standardized characteristic vector, inputs the standardized characteristic vector into a causal inference model, deduces the root cause and potential consequences of an abnormal state of a power grid, generates a decision result and a causal evidence chain, compresses the decision result and the causal evidence chain into an interpretable knowledge packet, extracts causal characteristic data, calculates causal characteristic data hash values, triggers a cooperative verification timer, verifies the integrity of the causal characteristic data, invokes historical causal characteristic data, generates signature feedback information, and performs pass rate verification on all signature feedback information before the cooperative verification timer, generates a cooperative verification instruction when the pass rate exceeds a preset pass threshold value, executes the cooperative verification module and performs a causal verification operation, and a causal agreement optimization result, and a master station performs a better-order, and a dynamic optimization packet, and a dynamic iteration packet is completed by the cooperative verification module. As a preferable scheme of the intelligent electric power fusion terminal, the electric quantity data comprise three-phase voltage data, three-phase current data and zero sequence voltage data; The temperature monitoring data comprise distribution transformer oil temperature data and terminal box body internal temperature data. As an optimal scheme of the intelligent power fusion terminal, the method comprises the following steps of generating a multi-mode operation data set, Performing feature extraction on three-phase voltage data, three-phase current data and zero sequence voltage data in the electric quantity data to generate an electric quantity feature vector; extracting characteristics of the temperature data of the distribution transformer oil in the temperature monitoring data and the temperature data in the terminal box body to generate a temperature characteristic vector; The sampling frequency of the electrical quantity characteristic vector is kept consistent with that of the temperature characteristic vector through interpolation processing, so that time synchronizatio