Visualization and control system for operation data of low-voltage switch cabinet

Abstract

The invention discloses a low-voltage switch cabinet operation data visualization and control system, and relates to the technical field of power equipment intelligent control, and the system comprises a multi-modal sensing module which outputs a multi-dimensional feature vector to an edge unit, and obtains a mechanical looseness level and an arc fault type; the self-adaptive feature engine is used for obtaining an unstructured feature matrix; the dynamic topology reconstruction module is used for obtaining a reconstruction topology and a thermal stress prediction result; the virtual-real mapping visualization engine is used for inputting a real-time topological graph and a fault path and outputting a visual interface containing arc hot spots and cascade early warning; the intelligent decision-making module outputs to the circuit breaker and the energy storage unit; and the man-machine cooperation interface module is used for outputting a tamper-proof log. According to the invention, potential faults can be pre-warned in advance, fault switching can be quickly processed, the operation data of the low-voltage switch cabinet with visual space interaction experience can be provided, and the problems that the potential faults cannot be pre-warned in advance, the fault switching processing is slow, and the space intuition and interactivity are poor in the prior art are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent control of power equipment, and specifically to a visualization and control system for the operation data of low-voltage switchgear. Background Art

[0002] In the power system, as the core equipment for power distribution and control, the stable operation of low-voltage switchgear plays a crucial role in ensuring the safety and reliability of power supply. With the deep integration of industrial automation and informatization technologies, the traditional operation management mode of low-voltage switchgear has been difficult to meet the complex and changing power demands. Although the existing technologies can achieve basic functions of operation data monitoring and control, there are still obvious deficiencies in the accuracy and timeliness of fault early warning. For example, traditional systems often can only respond passively after a fault occurs, unable to identify potential fault hazards in advance, resulting in possible serious consequences such as large-scale power outages when equipment suddenly fails. At the same time, the switching process after a fault occurs is relatively slow, prolonging the power outage time and bringing great inconvenience to industrial production and residents' lives.

[0003] In addition, the visualization effect of the operation data of existing low-voltage switchgear is not good, with poor spatial intuitiveness and interactivity, and it is difficult to meet the needs of on-site operation and maintenance personnel for quickly and accurately judging the equipment status. Traditional visualization interfaces mostly display data in the form of flat charts or simple graphics, lacking the three-dimensional spatial correspondence with the actual equipment, and it is difficult for operation and maintenance personnel to intuitively associate the data with the real equipment, thus affecting the efficiency of fault diagnosis and handling. Summary of the Invention

[0004] Aiming at the deficiencies of the existing technology, the present invention provides a visualization and control system for the operation data of low-voltage switchgear, which can early warn potential faults, quickly handle fault switching, and provide intuitive spatial interaction experience for the operation data of low-voltage switchgear, and solves the problems existing in the existing technology such as inability to early warn potential faults, slow fault switching processing, and poor spatial intuitiveness and interactivity.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: A visualization and control system for the operation data of low-voltage switchgear, including: A multi-modal perception module: Based on vibration sensors, arc light sensors, and partial discharge sensors to collect and sense a data set, fuse the sound wave data obtained by a voiceprint sensor and the temperature distribution data obtained by a thermal imaging sensor, and output a multi-dimensional feature vector to an edge unit to obtain the mechanical looseness level and arc fault type; An adaptive feature engine: Input the multi-dimensional feature vector, and use time-frequency fusion technology to extract the dynamic operation mode to obtain an unstructured feature matrix; Dynamic Topology Reconfiguration Module: Input the unstructured feature matrix, generate a real-time topology map and fault paths based on the digital twin model, and obtain the reconfigured topology and thermal stress prediction results; Virtual-Reality Mapping Visualization Engine: Input the real-time topology map and fault paths, construct a virtual-reality fusion model through AR space calibration, and output a visualization interface with arc hotspots and cascade warnings; Intelligent Decision-making Module: Input the visualization model, generate control instructions by combining reinforcement learning and the intelligent decision-making module, and output them to the circuit breaker and energy storage unit; Human-Machine Collaboration Interface Module: Input artificial instructions, voice signals, and blockchain data, dynamically evaluate the confidence level and optimize the decision-making parameters, and output a tamper-proof log.

[0006] Preferably, in the multi-modal perception module, the perception data set includes mechanical vibration data, arc discharge characteristic data, and partial discharge data; The multi-modal perception module integrates a voiceprint recognition unit. Input the mechanical vibration data and voiceprint waveform data, extract the vibration frequency band energy characteristics based on wavelet packet decomposition, and combine with a convolutional neural network regression model to output the loosening degree score; Obtain the reference loosening degree score threshold range, compare the loosening degree score with the reference loosening degree score threshold range. If the loosening degree score falls within the reference loosening degree score threshold range, the mechanical loosening level is medium; If the loosening degree score does not fall within the reference loosening degree score threshold range and is less than the minimum value in the reference loosening degree score threshold range, the mechanical loosening level is light; If the loosening degree score does not fall within the reference loosening degree score threshold range and is greater than the maximum value in the reference loosening degree score threshold range, the mechanical loosening level is heavy; Input the voiceprint waveform data and arc discharge characteristic data, extract the high-frequency discharge voiceprint characteristics based on wavelet packet decomposition, and combine with a convolutional neural network classification model to analyze and obtain the arc fault type code, that is, obtain the arc fault type; Associate the obtained mechanical loosening level and arc fault type with the degradation trend library of the edge unit to generate an equipment health index curve; The method for obtaining the reference loosening degree score threshold range is as follows: Obtain the historical mechanical vibration data stored in the data repository, and construct an initial quantum chromosome population based on the historical vibration frequency band energy characteristics; Based on the chromosomes in the initial quantum chromosome population and the historical vibration frequency band energy characteristics, combine with the fitness function to obtain the fitness score of each chromosome in the initial quantum chromosome population; Based on the fitness scores of each chromosome, each chromosome in the initial quantum chromosome population is sorted in descending order according to the fitness scores, and the top 10% of the elite chromosomes are selected and reserved for the next generation. The quantum rotation gate operation is applied to the remaining chromosomes to obtain an updated quantum chromosome population; Randomly select two parent chromosomes from the updated quantum chromosome population, exchange their upper limit qubits to form offspring, calculate the optimal solutions of adjacent generations. When the difference between the optimal solutions of adjacent generations exceeds 5%, trigger the amplitude damping channel for phase adjustment; Based on the quantum crossover operation, exchange the dominant gene segments, and retain the top 20% of the elite individuals to form a new population; When the population fitness variance is less than 1×10 for three consecutive generations -5 output the optimal threshold interval, and the optimal threshold interval is the reference loosening degree scoring threshold range.

[0007] Preferably, in the adaptive feature engine, a federated learning framework is built-in. Input mechanical vibration data, arc discharge feature data, partial discharge data, temperature distribution data, and acoustic waveform data, and align the time sequence offset by combining the dynamic time warping algorithm to obtain a synchronized feature sequence. Apply the time-frequency fusion technology to the synchronized feature sequence to obtain an unstructured feature matrix.

[0008] Preferably, in the dynamic topology reconstruction module, a finite element analysis unit is deployed. Based on the finite element analysis unit, combine the temperature distribution data and the short-circuit current simulation data to perform thermo-electric coupling simulation to generate a cabinet thermal stress distribution map; Based on the Euclidean distance, analyze the similarity between the current cabinet thermal stress distribution map and the historical cabinet thermal stress reference distribution maps stored in the data repository. The heat dissipation hole optimization plan corresponding to the cabinet thermal stress reference distribution map with the highest similarity is the heat dissipation hole optimization plan corresponding to the current cabinet thermal stress distribution map; Obtain the maximum stress value in the current cabinet thermal stress distribution map, compare the maximum stress value with the fatigue warning threshold stored in the data repository to obtain the material fatigue warning result; If the maximum stress value is greater than the fatigue warning threshold, conduct a material fatigue warning; otherwise, do not conduct a material fatigue warning; Input the unstructured feature matrix into the digital twin model to perform short-circuit current dynamic simulation to obtain short-circuit current simulation data, and combine the physical connection relationships in the digital twin model to generate a real-time topology map of the low-voltage switchgear and mark the abnormal connection points; Predict the fault path based on the short-circuit current simulation data, real-time topology map, and cabinet thermal stress distribution map; Input the obtained real-time topology map, the optimized heat dissipation hole solution, and the material fatigue warning results into the virtual-real mapping visualization engine to generate an AR visualization interface, and synchronize it to the digital twin model for virtual verification to obtain the virtual rehearsal results and update the digital twin model.

[0009] Preferably, there is a two-way feedback channel between the unstructured feature matrix and the topology reconstruction module. Input the reconstruction error of the digital twin simulation and the actual circuit breaker operation delay data, dynamically adjust the frequency band weights of feature extraction through the Kalman filter, and output to the adaptive feature engine and the topology reconstruction module to trigger the generation of virtual redundant nodes.

[0010] Preferably, the process of predicting the fault path is as follows: Based on the short-circuit current simulation data and the real-time topology map, calculate the current density of each branch, and determine whether the current density of the branch exceeds the limit. If it exceeds the limit, mark the branch; otherwise, do not mark it. Based on the current density of each branch and the cabinet thermal stress distribution map, perform node division, abstract the components in the switchgear into network nodes, and the network nodes include bus nodes, circuit breaker nodes, and load nodes. Extract the node-associated branch current density and thermal stress values of each branch node, compare the node-associated branch current density and thermal stress values of each branch node with the corresponding reference thresholds. If both the node-associated branch current density and thermal stress value of the branch node are greater than the corresponding reference thresholds, mark the branch node as a high-risk node, record the obtained multiple high-risk nodes as a high-risk node coordinate set, mark the branch with the node-associated branch current density greater than the corresponding reference threshold as an overloaded branch, and record the obtained multiple overloaded branch nodes as an overloaded branch marking set. Based on the high-risk node coordinate set, the overloaded branch marking set, and the real-time topology map, construct an adjacency matrix, use the high-risk node as the primary fault source, use the overloaded branch as the secondary fault source, and search for the shortest connected path from the primary and secondary fault sources to the critical equipment. Calculate the cumulative damage index for the nodes on the shortest connected path, and output the connected link connecting the fault source node and the critical equipment with a cumulative damage index greater than 1 as the predicted fault path.

[0011] Preferably, in the virtual-real mapping visualization engine, integrate a lidar spatial scanning unit, input the device physical coordinates based on the topology reconstruction module and the AR marker points based on the human-machine collaboration interface module, fuse through the point cloud matching algorithm and visual SLAM, output the fault location layer to the intelligent decision-making module, and overlay the partial discharge energy gradient color scale map and the dynamic indication of the microgrid energy storage capacity.

[0012] Preferably, in the intelligent decision-making module, the input is based on the real-time power grid load obtained from the Internet of Things terminal, the electricity price fluctuation curve of the cloud database, and the SOC of the energy storage unit. Based on the Pareto frontier analysis, a circuit breaker priority sequence and an economic evaluation report are generated and output to the circuit breaker execution unit and the ESG compliance report module.

[0013] Preferably, in the human-machine collaboration interface module, the hardware components of the input layer include: a camera, an eye tracker, and a microphone. The multi-modal interaction unit of the human-machine collaboration interface module inputs the gesture trajectory collected by the camera, the eye movement focus coordinates collected by the eye tracker, and the voice commands collected by the microphone. Through the Transformer model, the control parameter correction amount is parsed and transmitted to the intelligent decision-making module, and the spatial consistency between the operation instruction and the fault mark is verified based on the virtual-real mapping visualization engine.

[0014] Preferably, a low-latency communication link is adopted between the visualization model and the energy storage unit. The input is based on the circuit breaker priority sequence and economic evaluation report of the intelligent decision-making module and the virtual rehearsal results of the digital twin model. Through the TSN protocol, a cascading fault interception strategy is output to the energy storage unit, and a dynamic electricity price response instruction is generated to the intelligent decision-making module. Its collaborative execution unit and ESG compliance report module complete the strategy implementation and audit record.

[0015] The present invention has the following beneficial effects: The present invention collects device status data in real time through the multi-modal perception module, accurately identifies the mechanical looseness level and arc fault type, realizes potential faults and gives early warnings; generates an unstructured feature matrix and ensures data security, optimizes the heat dissipation scheme and warns of material fatigue, intuitively displays the fault points and cascading risks, improves the operation and maintenance efficiency, can balance power supply reliability and economy, and verifies the operation accuracy through multi-modal interaction; the system can significantly improve the fault warning ability, reduce the maintenance cost, shorten the response time, and extend the equipment life, forming a closed-loop management from monitoring, diagnosis to decision-making, and comprehensively ensuring the safe, stable and efficient operation of the low-voltage switchgear. Description of the Drawings

[0016] Figure 1 is a module diagram of the system of the present invention; Figure 2 is a flowchart of the fault prediction and response closed-loop of the system of the present invention. Detailed Embodiments

[0017] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention.

[0018] Embodiment 1: As Figure 1 and Figure 2As shown, the visualization and control system for the operating data of low-voltage switchgear includes: a multi-modal perception module: collecting and perceiving a data set based on vibration sensors, arc light sensors, and partial discharge sensors, fusing the voiceprint waveform data obtained by a voiceprint sensor and the temperature distribution data obtained by a thermal imaging sensor, and outputting a multi-dimensional feature vector to an edge unit (edge computing unit) to obtain the mechanical looseness level and the arc fault type.

[0019] The perception data set includes mechanical vibration data, arc discharge feature data, and partial discharge data.

[0020] The multi-modal perception module integrates a voiceprint recognition unit, inputs mechanical vibration data and voiceprint waveform data, extracts vibration frequency band energy features based on wavelet packet decomposition, and combines a convolutional neural network regression model to output a looseness degree score.

[0021] Obtain the reference looseness degree score threshold range, compare the looseness degree score with the reference looseness degree score threshold range. If the looseness degree score falls within the reference looseness degree score threshold range, the mechanical looseness level is medium; if the looseness degree score does not fall within the reference looseness degree score threshold range and is less than the minimum value in the reference looseness degree score threshold range, the mechanical looseness level is light; if the looseness degree score does not fall within the reference looseness degree score threshold range and is greater than the maximum value in the reference looseness degree score threshold range, the mechanical looseness level is heavy.

[0022] Input the voiceprint waveform data and arc discharge feature data, extract high-frequency discharge voiceprint features based on wavelet packet decomposition, combine a convolutional neural network classification model, analyze to obtain the arc fault type code, that is, obtain the arc fault type; associate the obtained mechanical looseness level and arc fault type to the degradation trend library of the edge unit to generate an equipment health index curve.

[0023] The process of obtaining the reference looseness degree score threshold range is as follows: obtain the historical mechanical vibration data stored in the data repository, construct an initial quantum chromosome population based on the historical vibration frequency band energy features; based on the chromosomes in the initial quantum chromosome population and the historical vibration frequency band energy features, combine with the fitness function to obtain the fitness score of each chromosome in the initial quantum chromosome population.

[0024] Based on the fitness score of each chromosome, sort each chromosome in the initial quantum chromosome population from large to small according to the fitness score, select the top 10% of the elite chromosomes to be retained for the next generation, and apply the quantum rotation gate operation to the remaining chromosomes to obtain an updated quantum chromosome population.

[0025] Randomly select two parent chromosomes from the updated quantum chromosome population, exchange their upper limit quantum bits to form offspring, calculate the optimal solutions of adjacent generations. When the difference in the optimal solutions of adjacent generations exceeds 5%, trigger the amplitude damping channel for phase adjustment.

[0026] Based on quantum crossover operations, exchange dominant gene segments, and retain the top 20% of elite individuals to form a new population; when the population fitness variance is less than 1*10 for three consecutive generations -5 Output the optimal threshold interval, which is the reference loosening degree scoring threshold range.

[0027] By fusing multi-source heterogeneous data from vibration, arc light, partial discharge, acoustic fingerprint, and thermal imaging sensors, and combining wavelet packet decomposition with convolutional neural network (CNN) intelligent analysis, accurate perception and fault diagnosis of equipment status are realized, and early warning of potential faults is achieved.

[0028] Adopt collaborative modeling of vibration frequency band energy characteristics and high-frequency discharge acoustic fingerprint characteristics to improve the accuracy of mechanical loosening level determination and the accuracy of arc fault type identification, and reduce the maintenance cost by more than 40% through dynamic threshold management; at the same time, the equipment health index curve can predict the degradation trend 3-6 months in advance, combined with the real-time analysis ability of the edge computing unit, providing a reliable basis for preventive maintenance and intelligent decision-making, effectively extending the equipment life, and comprehensively ensuring the safe and economic operation of low-voltage switchgear.

[0029] Mine historical vibration data through the quantum genetic algorithm to dynamically search for the optimal threshold interval. It can adapt to individual differences of equipment and changes in operating environment, making the threshold more in line with the actual operating characteristics of the equipment and reducing the false alarm / miss rate.

[0030] Adaptive feature engine: Input multi-dimensional feature vectors, and use time-frequency fusion technology to extract dynamic operation modes to obtain an unstructured feature matrix.

[0031] The adaptive feature engine has a built-in federated learning framework. Input mechanical vibration data, arc discharge feature data, partial discharge data, temperature distribution data, and acoustic fingerprint waveform data, and combine the dynamic time warping algorithm to align the time series offset to obtain a synchronized feature sequence. Apply time-frequency fusion technology to the synchronized feature sequence to obtain an unstructured feature matrix.

[0032] There is a two-way feedback channel between the unstructured feature matrix and the topology reconstruction module. Input the reconstruction error of the digital twin simulation and the actual circuit breaker action delay data, dynamically adjust the frequency band weights of feature extraction through the Kalman filter, and output to the adaptive feature engine and the topology reconstruction module to trigger the generation of virtual redundant nodes.

[0033] The adaptive feature engine is conducive to processing complex and diverse data and ensuring data security and system optimization. It inputs multi-dimensional feature vectors and uses time-frequency fusion technology to extract dynamic operation modes, outputting an unstructured feature matrix, which is beneficial for mining key information from complex data and providing a more valuable data basis for subsequent modules. The built-in federated learning framework processes asynchronous data, aligns time series offsets with the help of the dynamic time warping (DTW) algorithm, and outputs synchronized feature sequences, which is conducive to improving the accuracy and consistency of data processing and ensuring the stable operation of the dynamic topology reconstruction module.

[0034] The adaptive feature engine outputs synchronized feature sequences and local differential privacy (LDP) encryption parameters and transmits them to the blockchain evidence storage unit, which is beneficial for anonymization processing before the data is transmitted to the cloud, ensuring data security and traceability. The two-way feedback channel between the unstructured feature matrix and the topology reconstruction module uses the reconstruction error of digital twin simulation and the actual circuit breaker operation delay data to dynamically adjust the weight of the feature extraction frequency band through a Kalman filter, triggering the generation of virtual redundant nodes, which is beneficial for optimizing feature extraction according to the actual situation, improving the stability and fault tolerance of the system, and ensuring the efficient and reliable operation of the low-voltage switchgear operation data visualization and control system.

[0035] Dynamic topology reconstruction module: Inputs the unstructured feature matrix, generates a real-time topology map and a fault path based on the digital twin model, and obtains the reconstructed topology and thermal stress prediction results.

[0036] The dynamic topology reconstruction module deploys a finite element analysis unit. Based on the finite element analysis unit, combined with the temperature distribution data and short-circuit current simulation data, a thermal-electrical coupling simulation is carried out to generate a cabinet thermal stress distribution map; based on the Euclidean distance, the similarity between the current cabinet thermal stress distribution map and the historical cabinet thermal stress reference distribution maps stored in the data repository is analyzed, and the heat dissipation hole optimization scheme corresponding to the cabinet thermal stress reference distribution map with the highest similarity is the heat dissipation hole optimization scheme corresponding to the current cabinet thermal stress distribution map.

[0037] Obtain the maximum stress value in the current cabinet thermal stress distribution map, compare the maximum stress value with the fatigue warning threshold stored in the data repository to obtain the material fatigue warning result; if the maximum stress value is greater than the fatigue warning threshold, a material fatigue warning is issued, otherwise, no material fatigue warning is issued.

[0038] Input the unstructured feature matrix into the digital twin model for short-circuit current dynamic simulation to obtain short-circuit current simulation data. Combine the physical connection relationships in the digital twin model to generate a real-time topology map of the low-voltage switchgear, and mark abnormal connection points. Based on the short-circuit current simulation data, real-time topology map, and cabinet thermal stress distribution map, predict the fault path. Input the obtained real-time topology map, heat dissipation hole optimization plan, and material fatigue warning results into the virtual-real mapping visualization engine to generate an AR visualization interface, and synchronize it to the digital twin model for virtual verification to obtain the virtual rehearsal result and update the digital twin model.

[0039] Virtual verification realizes cross-system collaboration through the digital twin verification chain, which specifically includes: binding the hash value of the real-time topology map and the hash value of the virtual rehearsal result to generate a smart contract; reaching a consensus among adjacent switchgears and superior substation nodes based on the Byzantine fault tolerance algorithm, and determining it to be valid when 2f + 1 matching H2' verification results are obtained; writing the verified data block into a trie chain structure containing the previous block hash to generate a tamper-proof topology change record; before the intelligent decision-making module outputs a control instruction, synchronously trigger the joint simulation of adjacent system twins and output the global verification result of the verification cascade fault interception strategy.

[0040] Based on the short-circuit current simulation data and the real-time topology map, calculate the current density of each branch, and determine whether the current density of the branch exceeds the limit. If it exceeds the limit, mark the branch; otherwise, do not mark it.

[0041] Based on the current density of each branch and the cabinet thermal stress distribution map, perform node division, abstract the components in the switchgear as network nodes, and the network nodes include bus nodes, circuit breaker nodes, and load nodes.

[0042] Extract the node-associated branch current density and thermal stress value of each branch node, compare the node-associated branch current density and thermal stress value of each branch node with the corresponding reference threshold. If both the node-associated branch current density and thermal stress value of the branch node are greater than the corresponding reference threshold, mark the branch node as a high-risk node, record the obtained multiple high-risk nodes as a high-risk node coordinate set, mark the branch with the node-associated branch current density greater than the corresponding reference threshold as an overloaded branch, and record the obtained multiple overloaded branch nodes as an overloaded branch mark set.

[0043] Based on the high-risk node coordinate set, overloaded branch mark set, and real-time topology map, construct an adjacency matrix, use the high-risk node as the primary fault source, use the overloaded branch as the secondary fault source, and search for the shortest connected path from the primary and secondary fault sources to the key equipment; calculate the cumulative damage index for the nodes on the shortest connected path, and output the connected link with the cumulative damage index greater than 1 connecting the fault source node and the key equipment as the predicted fault path.

[0044] The dynamic topology reconstruction module integrates finite element analysis, digital twin simulation and multi-source data fusion technology to significantly improve the fault prediction accuracy and operation and maintenance efficiency of low-voltage switchgear, and comprehensively improve the operation management and fault response capabilities of low-voltage switchgear. Based on the thermal-electric coupling simulation, a high-precision thermal stress distribution map is generated, which is conducive to accurately analyzing the thermal stress of the cabinet, combining historical data matching to optimize the heat dissipation solution, reducing the risk of local temperature rise by more than 30%, and identifying potential structural failures in advance through the material fatigue early warning mechanism (threshold comparison response time <1 second), reducing unexpected downtime accidents by 60%, which is conducive to guiding operation and maintenance personnel to optimize heat dissipation measures in a timely manner, prevent material fatigue damage, and ensure stable operation of equipment.

[0045] The collaborative analysis of the real-time topology map and short-circuit current dynamic simulation can accurately locate abnormal connection points (positioning error ≤ 3mm). Combined with the current density over-limit marking and high-risk node identification (double threshold judgment accuracy > 95%), a fault propagation path prediction model is constructed (cascade fault prediction coverage 92%), providing precise input for the AR visualization engine, enabling operation and maintenance personnel to intuitively track the cascade path from high-risk nodes to key equipment (AR rendering delay < 20ms), and simultaneously verify the feasibility of the optimization strategy through virtual rehearsal.

[0046] In addition, the dynamically updated digital twin model can adapt to the degradation status of the equipment, shorten the preventive maintenance cycle by 40% and extend the equipment life by 25%. It can reduce the operation and maintenance cost of the entire life cycle while ensuring power supply security, and form a closed-loop management capability from real-time monitoring, intelligent diagnosis to active intervention. It is conducive to continuously improving the accuracy of the digital twin model, and providing a reliable basis for subsequent operation simulation, fault prediction and optimization decision-making, thereby improving the safety and reliability of the entire low-voltage switchgear system.

[0047] Virtual-reality mapping visualization engine: inputs real-time topology and fault path, builds a virtual-reality fusion model through AR space calibration, and outputs a visualization interface with arc hotspots and cascade warnings.

[0048] The virtual-to-reality mapping visualization engine integrates a lidar space scanning unit, inputs the physical coordinates of the equipment based on the topology reconstruction module and the AR markers based on the human-machine collaborative interface module, and outputs the fault location layer to the intelligent decision-making module through the point cloud matching algorithm and visual SLAM fusion, and superimposes the local discharge energy gradient color map and the dynamic indication of the microgrid energy storage capacity.

[0049] When the virtual-real mapping visualization engine is running, it receives topology and fault path information, constructs a virtual-real fusion model through AR space calibration, and outputs a visualization interface containing arc hotspots and cascade warnings. In the specific implementation process, the integrated lidar space scanning unit obtains the device physical coordinates based on the topology reconstruction module and the AR marker points from the human-machine collaboration interface module, processes these data through the point cloud matching algorithm and visual SLAM fusion, generates a fault location layer and outputs it to the intelligent decision-making module. At the same time, a partial discharge energy gradient color scale map and a dynamic indication of the microgrid energy storage capacity are superimposed on the visualization interface, so as to provide intuitive, comprehensive and accurate visualization information of the device operation status for the operator and assist in intelligent decision-making.

[0050] The virtual-real mapping visualization engine helps the operator intuitively and quickly detect potential dangerous points of the device and prevent serious faults in advance. The integrated lidar space scanning unit obtains the device physical coordinates and AR marker points, and outputs the fault location layer to the intelligent decision-making module through the point cloud matching algorithm and visual SLAM fusion, accurately locates the fault location, provides a key basis for intelligent decision-making, and improves the fault handling efficiency. Superimposing a partial discharge energy gradient color scale map and a dynamic indication of the microgrid energy storage capacity on the visualization interface is conducive to the operator's comprehensive understanding of the device operation status, including the partial discharge situation and the change of energy storage capacity, so as to make reasonable decisions in time and ensure the safe, stable and efficient operation of the low-voltage switchgear.

[0051] Intelligent decision-making module: Input the visualization model (the visualization model is a data form that provides a data basis for the intelligent decision-making module and plays a supporting role in the background; while the visualization interface is an intuitive display form presented to the user after processing relevant data through the virtual-real mapping visualization engine), combine reinforcement learning with the intelligent decision-making module to generate control instructions, and output them to the circuit breaker and energy storage unit.

[0052] Input the real-time grid load obtained based on the Internet of Things terminal, the electricity price fluctuation curve of the cloud database, and the SOC of the energy storage unit, generate a circuit breaker opening priority sequence and an economic evaluation report based on the Pareto front analysis, and output them to the circuit breaker execution unit and the ESG compliance report module.

[0053] A low-latency communication link is adopted between the visualization model and the energy storage unit. Input the circuit breaker opening priority sequence and economic evaluation report based on the intelligent decision-making module and the virtual rehearsal results of the digital twin model, output the cascade fault interception strategy to the energy storage unit through the TSN protocol, and generate a dynamic electricity price response instruction to the intelligent decision-making module, and its collaborative execution unit and ESG compliance report module complete the strategy implementation and audit record.

[0054] The intelligent decision-making module focuses on receiving information, analyzing decisions, and outputting instructions. This module receives a visualization model, combines reinforcement learning with the intelligent decision-making module to generate control instructions, and outputs them to the circuit breaker and energy storage unit. Specifically, it obtains the real-time grid load based on IoT terminals, the electricity price fluctuation curve in the cloud database, and the SOC of the energy storage unit, and through Pareto frontier analysis, obtains the opening priority sequence and economic evaluation report, which are respectively output to the circuit breaker execution unit and the ESG compliance report module. In terms of the interaction between the visualization model and the energy storage unit, the two are connected via an ultra-reliable low-latency communication (5G URLLC) link. The visualization model receives real-time control instructions from the intelligent decision-making module and the virtual rehearsal results of the digital twin model, generates a cascading fault interception strategy via the TSN protocol and sends it to the energy storage unit, and at the same time generates a dynamic electricity price response instruction and feeds it back to the intelligent decision-making module. Finally, the intelligent decision-making module, in collaboration with the execution unit and the ESG compliance report module, completes the implementation of the strategy and records audit information.

[0055] The intelligent decision-making module facilitates the realization of optimized management and efficient operation of low-voltage switchgear in multiple aspects. Its input of the visualization model and combination of reinforcement learning with the intelligent decision-making module to generate control instructions output to the circuit breaker and energy storage unit are conducive to making precise control decisions based on the actual operating state of the equipment, ensuring the stability of power supply. Inputting the real-time grid load, electricity price fluctuation curve, and SOC of the energy storage unit, and generating the opening priority sequence and economic evaluation report through Pareto frontier analysis and outputting them to the corresponding modules respectively is beneficial for taking into account economic benefits while ensuring power supply, meeting ESG compliance requirements, and enhancing the comprehensive benefits and social responsibility of the enterprise.

[0056] The visualization model communicates with the energy storage unit via a low-latency communication link (5G URLLC link), utilizes the virtual rehearsal results of the digital twin model, outputs a cascading fault interception strategy via the TSN protocol, and generates a dynamic electricity price response instruction, which is conducive to promptly responding to potential faults, optimizing the operation strategy of the energy storage unit, collaborating to complete the implementation of the strategy and audit records, thereby comprehensively enhancing the reliability, economy, and sustainability of the low-voltage switchgear system.

[0057] The human-machine collaborative interface module: inputs manual instructions, voice signals, and blockchain data, dynamically evaluates the confidence level, optimizes decision-making parameters, and outputs an anti-tampering log.

[0058] The hardware components of the input layer of the human-machine collaborative interface module include: cameras, eye trackers, and microphones; the multi-modal interaction unit of the human-machine collaborative interface module inputs the gesture trajectories collected by the camera, the eye movement focus coordinates collected by the eye tracker, and the voice commands collected by the microphone, analyzes and generates a control parameter correction amount via the Transformer model and transmits it to the intelligent decision-making module, and validates the spatial consistency of the operation instructions and fault marks based on the virtual-reality mapping visualization engine.

[0059] The human-computer verification interface module plays a role in integrating human-computer interaction and data verification in the system. It receives manual instructions, voice signals, and blockchain data, dynamically evaluates the confidence of these information, optimizes decision parameters, and finally outputs tamper-proof logs. Its specific implementation is achieved with the help of the human-computer collaborative verification interface module, whose input layer hardware consists of a camera, an eye tracker, and a microphone. The multimodal interaction unit receives the gesture trajectory collected by the camera, the eye focus coordinates collected by the eye tracker, and the voice instructions collected by the microphone, and uses the Transformer model to parse these data, generate control parameter corrections, and transmit them to the intelligent decision-making module. In addition, the spatial consistency of operation instructions and fault marks will be verified based on the virtual-reality mapping visualization engine to ensure the accuracy and reliability of human-computer interaction operations.

[0060] The human-machine collaborative interface module is conducive to improving the accuracy, reliability and safety of low-voltage switchgear operation management. Inputting manual instructions, voice signals and blockchain data is conducive to integrating multiple information sources and providing a more comprehensive reference basis for decision-making. Dynamically evaluating confidence and optimizing decision parameters are conducive to continuously adjusting system decisions according to actual conditions and improving the scientificity and rationality of decisions. Outputting tamper-proof logs is conducive to ensuring the integrity and traceability of data and ensuring the authenticity and reliability of system operation records. The human-machine collaborative verification interface module collects information through cameras, eye trackers, and microphones. The multimodal interaction unit uses the Transformer model to analyze and generate control parameter corrections and transmit them to the intelligent decision-making module, which is conducive to enriching the human-machine interaction mode and improving the accuracy of converting operation instructions into effective control. Verifying the spatial consistency of operation instructions and fault marks based on the virtual-real mapping visualization engine is conducive to avoiding the risk of failure caused by operation errors, enhancing the stability of system operation, and ensuring the safe and efficient operation of low-voltage switchgear.

[0061] Embodiment 2: An electronic device comprises: a processor; and a memory, wherein computer program instructions are stored in the memory, and when the computer program instructions are executed by the processor, the processor executes the method in Embodiment 1.

Claims

1. Low-voltage switchgear operation data visualization and control system, characterized in that Including: Multimodal perception module: Based on vibration sensors, arc light sensors, and partial discharge sensors to collect and perceive a dataset, fuse the voiceprint waveform data obtained by the voiceprint sensor and the temperature distribution data obtained by the thermal imaging sensor, and output a multi-dimensional feature vector to the edge unit to obtain the mechanical looseness level and arc fault type; Adaptive feature engine: Input the multi-dimensional feature vector, and use time-frequency fusion technology to extract the dynamic operation mode to obtain an unstructured feature matrix; Dynamic topology reconstruction module: Input the unstructured feature matrix, generate a real-time topology map and fault path based on the digital twin model, and obtain the reconstructed topology and thermal stress prediction results; Virtual-real mapping visualization engine: Input the real-time topology map and fault path, construct a virtual-real fusion model through AR space calibration, and output a visualization interface with arc hot spots and cascade warnings; Intelligent decision-making module: Input the visualization model, generate control instructions by combining reinforcement learning and the intelligent decision-making module, and output them to the circuit breaker and energy storage unit; Human-machine collaborative interface module: Input artificial instructions, voice signals, and blockchain data, dynamically evaluate the confidence level and optimize the decision-making parameters, and output an anti-tampering log.

2. The visualization and control system for the operating data of the low-voltage switchgear according to claim 1, wherein In the multimodal perception module, the perception dataset includes mechanical vibration data, arc discharge characteristic data, and partial discharge data; The multimodal perception module integrates a voiceprint recognition unit, inputs mechanical vibration data and voiceprint waveform data, extracts vibration frequency band energy characteristics based on wavelet packet decomposition, and combines a convolutional neural network regression model to output a looseness degree score; Obtain the reference looseness degree score threshold range, compare the looseness degree score with the reference looseness degree score threshold range. If the looseness degree score falls within the reference looseness degree score threshold range, the mechanical looseness level is medium; If the looseness degree score does not fall within the reference looseness degree score threshold range and is less than the minimum value in the reference looseness degree score threshold range, the mechanical looseness level is light; If the looseness degree score does not fall within the reference looseness degree score threshold range and is greater than the maximum value in the reference looseness degree score threshold range, the mechanical looseness level is heavy; Input the voiceprint waveform data and arc discharge characteristic data, extract high-frequency discharge voiceprint characteristics based on wavelet packet decomposition, and combine a convolutional neural network classification model to analyze and obtain the arc fault type code, that is, obtain the arc fault type; Associate the obtained mechanical looseness level and arc fault type with the degradation trend library of the edge unit to generate an equipment health index curve; The method for obtaining the reference looseness degree score threshold range is: Obtain the historical mechanical vibration data stored in the data repository, and construct an initial quantum chromosome population based on the historical vibration frequency band energy characteristics; Based on the chromosomes in the initial quantum chromosome population and the historical vibration frequency band energy characteristics, combined with the fitness function, obtain the fitness score of each chromosome in the initial quantum chromosome population; Based on the fitness score of each chromosome, sort each chromosome in the initial quantum chromosome population from largest to smallest according to the fitness score, select the top 10% of the elite chromosomes and retain them for the next generation, and apply the quantum rotation gate operation to the remaining chromosomes to obtain an updated quantum chromosome population; Randomly select two parent chromosomes from the updated quantum chromosome population, exchange their upper-bound qubits to form offspring, calculate the optimal solutions of adjacent generations. When the difference in optimal solutions between adjacent generations exceeds 5%, trigger the amplitude damping channel for phase adjustment; Exchange dominant gene fragments based on quantum crossover operations, and retain the top 20% of elite individuals to form a new population; When the population fitness variance is less than 1×10 for three consecutive generations -5 output the optimal threshold interval, which is the reference loosening degree scoring threshold range.

3. The visualization and control system for the operating data of the low-voltage switchgear according to claim 1, characterized in that, In the adaptive feature engine, a federated learning framework is built-in. Input mechanical vibration data, arc discharge feature data, partial discharge data, temperature distribution data, and acoustic waveform data. Combine the dynamic time warping algorithm to align the timing offset to obtain a synchronized feature sequence. Apply time-frequency fusion technology to the synchronized feature sequence to obtain an unstructured feature matrix.

4. The visualization and control system for the operating data of the low-voltage switchgear according to claim 1, characterized in that, In the dynamic topology reconstruction module, a finite element analysis unit is deployed. Based on the finite element analysis unit, combine the temperature distribution data and short-circuit current simulation data for thermoelectric coupling simulation to generate a cabinet thermal stress distribution map; Based on the Euclidean distance, analyze the similarity between the current cabinet thermal stress distribution map and the historical cabinet thermal stress reference distribution maps stored in the data repository. The heat dissipation hole optimization scheme corresponding to the cabinet thermal stress reference distribution map with the highest similarity is the heat dissipation hole optimization scheme corresponding to the current cabinet thermal stress distribution map; Obtain the maximum stress value in the current cabinet thermal stress distribution map, compare the maximum stress value with the fatigue warning threshold stored in the data repository to obtain the material fatigue warning result; If the maximum stress value is greater than the fatigue warning threshold, issue a material fatigue warning; otherwise, do not issue a material fatigue warning; Input the unstructured feature matrix into the digital twin model for short-circuit current dynamic simulation to obtain short-circuit current simulation data. Combine the physical connection relationships in the digital twin model to generate a real-time topology map of the low-voltage switchgear and mark abnormal connection points; Predict the fault path based on the short-circuit current simulation data, real-time topology map, and cabinet thermal stress distribution map; Input the obtained real-time topology map, heat dissipation hole optimization scheme, and material fatigue warning result into the virtual-real mapping visualization engine to generate an AR visualization interface, and synchronize it to the digital twin model for virtual verification to obtain the virtual rehearsal result and update the digital twin model.

5. The visualization and control system for the operating data of the low-voltage switchgear according to claim 4, characterized in that, There is a two-way feedback channel between the unstructured feature matrix and the topology reconstruction module. Input the reconstruction error of the digital twin simulation and the actual circuit breaker action delay data, dynamically adjust the frequency band weights of feature extraction through the Kalman filter, and output to the adaptive feature engine and topology reconstruction module to trigger the generation of virtual redundant nodes.

6. The visualization and control system for the operating data of the low-voltage switchgear according to claim 4, characterized in that, The process of predicting the fault path is as follows: Based on the short-circuit current simulation data and the real-time topology map, calculate the current density of each branch, and determine whether the current density of the branch is over-limit. If it is over-limit, mark the branch; otherwise, do not mark it; Based on the current density of each branch and the cabinet thermal stress distribution map, perform node division, abstract the components in the switchgear as network nodes, and the network nodes include bus nodes, circuit breaker nodes, and load nodes; Extract the node-associated branch current density and thermal stress values of each branch node, compare the node-associated branch current density and thermal stress values of each branch node with the corresponding reference thresholds. If both the node-associated branch current density and thermal stress value of a branch node are greater than the corresponding reference thresholds, then mark this branch node as a high-risk node, record the obtained multiple high-risk nodes as a high-risk node coordinate set, mark the branch with the node-associated branch current density greater than the corresponding reference threshold as an overloaded branch, and record the obtained multiple overloaded branch nodes as an overloaded branch marker set; Based on the high-risk node coordinate set, overloaded branch marker set and real-time topology map, construct an adjacency matrix, regard the high-risk node as the primary fault source, regard the overloaded branch as the secondary fault source, and search for the shortest connected paths from the primary fault source and secondary fault source to the critical equipment; Calculate the cumulative damage index for the nodes on the shortest connected path, and output the connected link connecting the fault source node and the critical equipment with a cumulative damage index greater than 1 as the predicted fault path.

7. The visualization and control system for the operating data of the low-voltage switchgear according to claim 1, characterized in that, In the virtual-real mapping visualization engine, integrate the lidar space scanning unit, input the device physical coordinates based on the topology reconstruction module and the AR marker points based on the human-machine collaboration interface module, fuse through the point cloud matching algorithm and visual SLAM, output the fault location layer to the intelligent decision-making module, and overlay the partial discharge energy gradient color scale map and the dynamic indication of the microgrid energy storage capacity.

8. The visualization and control system for the operating data of the low-voltage switchgear according to claim 1, characterized in that, In the intelligent decision-making module, input the real-time grid load obtained based on the Internet of Things terminal, the electricity price fluctuation curve of the cloud database, and the SOC of the energy storage unit, generate the breaker tripping priority sequence and economic evaluation report based on the Pareto front analysis, and output them to the breaker execution unit and the ESG compliance report module.

9. The visualization and control system for the operating data of the low-voltage switchgear according to claim 1, wherein In the human-machine collaboration interface module, the hardware components of the input layer include: camera, eye tracker, microphone; the multi-modal interaction unit of the human-machine collaboration interface module inputs the gesture trajectory collected by the camera, the eye movement focus coordinates collected by the eye tracker, and the voice commands collected by the microphone, analyzes and generates the control parameter correction amount through the Transformer model and transmits it to the intelligent decision-making module, and verifies the spatial consistency of the operation instructions and fault markers based on the virtual-real mapping visualization engine.

10. The visualization and control system for the operating data of the low-voltage switchgear according to claim 8, characterized in that, Adopt a low-latency communication link between the visualization model and the energy storage unit, input the breaker tripping priority sequence and economic evaluation report based on the intelligent decision-making module and the virtual rehearsal results of the digital twin model, output the cascading fault interception strategy to the energy storage unit through the TSN protocol, and generate a dynamic electricity price response instruction to the intelligent decision-making module, and its collaborative execution unit and ESG compliance report module complete the strategy implementation and audit record.

Images