Substation line fault diagnosis processing system and method using artificial intelligence

Abstract

The invention discloses a transformer substation line fault diagnosis processing system and method using artificial intelligence. The system comprises a monitoring module, a diagnosis module and a processing module. The monitoring module comprises a distributed sensor node, a quantum sensor and an environment adaptability subunit, can collect line temperature, electricity and voltage data and is adaptive to the environment. The diagnosis module comprises an artificial intelligence algorithm unit and can analyze data to determine fault types and positions and update parameters online. The processing module comprises an automatic control unit and a virtual repair simulation subunit, and the virtual repair simulation subunit can simulate a fault injection and repair operation evaluation scheme. The method comprises the steps of data acquisition, fault analysis, virtual repair simulation, actual fault isolation repair and the like. According to the method, the line state can be comprehensively monitored, the fault can be accurately diagnosed, the fault processing reliability is improved, the repair process is simulated in the virtual environment, unnecessary losses and risks caused in the actual repair process are avoided, and the fault processing reliability is improved.

Description

Technical Field

[0001] The present invention relates to the field of substation line fault diagnosis and processing applications, and particularly to a substation line fault diagnosis and processing system and method using artificial intelligence. Background Art

[0002] With the rapid development of the new energy photovoltaic power generation industry, as a key node in the power system, the efficiency of substation line fault diagnosis and processing directly affects the stability and reliability of the power grid. At present, the scale of new energy photovoltaic power stations is constantly expanding, and the complexity and failure rate of substation lines are also increasing. Traditional fault diagnosis methods mainly rely on manual inspections and experience judgments, which are inefficient and prone to missed detections. In recent years, the application of artificial intelligence technology in the power system has gradually increased, but there are still many challenges in substation line fault diagnosis and processing, such as insufficient real-time performance, low diagnostic accuracy, and cumbersome processing procedures.

[0003] Common substation line fault diagnosis and processing mainly adopt the following methods: First is the method based on manual inspection: By regularly inspecting line equipment, recording abnormal situations and processing them. The advantage is intuitive and reliable, and the disadvantages are low efficiency, high cost, and difficulty in covering all fault points. Second is the method based on sensor monitoring: Installing sensors such as temperature, current, and voltage at key nodes of the line to monitor the line status in real time. The advantage is that partial automated monitoring can be achieved, and the disadvantage is that the number of sensors is limited and it is difficult to comprehensively cover complex lines. Finally is the fault diagnosis method based on traditional algorithms: Using signal processing or statistical analysis algorithms to analyze monitoring data to identify faults. The advantage is a relatively high degree of automation, and the disadvantages are insufficient diagnostic accuracy for complex faults and poor algorithm adaptability, which cannot meet the working requirements of substation line fault diagnosis and processing applications. Therefore, a substation line fault diagnosis and processing system and method using artificial intelligence are proposed. Summary of the Invention

[0004] The present invention provides the following technical solution: A substation line fault diagnosis and processing system using artificial intelligence, comprising: A monitoring module, a diagnosis module, and a processing module. The monitoring module includes distributed sensor nodes and quantum sensors. The monitoring module is installed at the positions to be monitored on the substation line. The monitoring module is used to collect data on the temperature, current, and voltage of the line in real time. An environmental adaptability sub-unit is additionally provided inside the monitoring module. Miniature energy harvesting devices are integrated inside both the distributed sensor nodes and the quantum sensors. The miniature energy harvesting devices are coupled to the substation electromagnetic field; The diagnostic module is internally provided with an artificial intelligence algorithm unit. The diagnostic module is used to analyze the data collected by the monitoring module, identify the fault type and location. The artificial intelligence algorithm unit is provided with a graph neural network and a physical knowledge embedding model. The diagnostic module is internally built with an online learning function, which can dynamically update the model parameters according to new fault cases; The processing module is internally equipped with an automatic control unit. The processing module is used to generate processing instructions according to the diagnostic results and perform fault isolation and repair operations. The processing module is internally equipped with a virtual repair simulation sub-unit. The virtual repair simulation sub-unit is used to simulate the repair process in a virtual environment before performing the actual fault isolation and repair operations, and evaluate the feasibility and the resulting impact of the repair plan according to the simulation results.

[0005] The present invention provides a substation line fault diagnosis and processing method using artificial intelligence, including the following steps: S1 Monitoring module data acquisition step: First, the temperature sensors and Hall effect current sensors in the distributed sensor nodes monitor the substation line, and at the same time, the micro sensors in the environmental adaptability sub-unit collect the environmental parameters inside the substation. Subsequently, the intelligent control chip adjusts the working states of the distributed sensor nodes and the quantum sensors according to the collected environmental parameters, and finally transmits the collected data to the diagnostic module; S2 Diagnostic module fault analysis step: First, after receiving the data collected in step S1, the diagnostic module performs preliminary preprocessing on the received data; S21 Graph neural network analysis: Subsequently, the nodes and edges in the graph neural network perform operations according to the received data, and the data is weighted according to the actual situation of the line through the weight matrix of each layer, and the characteristic information in the data is analyzed to preliminarily judge the characteristics related to the fault area and fault type; S22 Physical knowledge embedding model analysis: The data processed in step S21 enters the physical knowledge embedding model and is matched at the bottom layer of basic physical parameters layer. Subsequently, the physical process in the line is simulated in the middle layer of physical process simulation layer according to the basic physical parameters to further analyze the fault type and location; S23 Comprehensive diagnosis and model update: The diagnostic module synthesizes the analysis results of the graph neural network and the physical knowledge embedding model to identify the fault type and location. When new fault cases occur, the online learning function of the diagnostic module will dynamically update the model parameters according to the new cases; S3 Processing module fault handling step: S31 Virtual repair simulation: After receiving the data in step S2, the virtual repair simulation subunit in the processing module is started, so that the fault injection module injects corresponding faults into the virtual model according to the fault type and location obtained by the diagnosis module. Subsequently, the repair operation simulation module follows the processing instructions generated by the processing module according to the diagnosis result and performs repair operation simulation in the virtual model; S32 Actual fault isolation and repair: When the result of the virtual repair simulation shows that the repair plan is feasible, the automatic control unit in the processing module generates processing instructions according to the diagnosis result and performs fault isolation and repair operations.

[0006] Preferably, the distributed sensor node includes a temperature sensor and a Hall effect current sensor. Inductive probes are installed outside both the temperature sensor and the Hall effect current sensor, and the inductive probes are distributed in an array.

[0007] Preferably, the quantum sensor is covered with a protective shell with electromagnetic shielding function. The outer layer of the protective shell with electromagnetic shielding function is a metal shielding layer, the middle layer of the protective shell with electromagnetic shielding function is an insulating buffer layer, and the inner layer of the protective shell with electromagnetic shielding function is a heat-conducting layer.

[0008] Preferably, the environmental adaptability subunit includes a micro sensor and an intelligent control chip, and the intelligent control chip is connected to the distributed sensor node and the quantum sensor.

[0009] Preferably, the nodes in the graph neural network represent various components and monitoring points in the substation line, the edges in the graph neural network represent the electrical connection relationships and physical associations between these components and monitoring points, and each layer of the graph neural network is provided with a weight matrix.

[0010] Preferably, the structure of the physical knowledge embedding model in the artificial intelligence algorithm unit is set as a hierarchical structure. The bottom layer of the physical knowledge embedding model is the basic physical parameter layer, and the basic physical parameter layer contains the basic physical characteristic parameters of the substation line. The middle layer of the physical knowledge embedding model is the physical process simulation layer.

[0011] Preferably, the virtual repair simulation subunit includes a virtual substation line model construction module, a fault injection module, and a repair operation simulation module. The virtual substation line model construction module is used to construct a virtual substation line model according to the actual layout of the substation, the line connection relationship, and the device parameter information. The fault injection module is used to inject corresponding faults into the virtual model according to the fault type and location obtained by the diagnosis module. The repair operation simulation module is used to perform repair operation simulation in the virtual model according to the processing instructions generated by the processing module.

[0012] Preferably, the temperature sensor and the Hall effect current sensor are installed at key nodes of the substation line, and the key nodes are determined according to the historical fault data, electrical topology, and heat exchange characteristics of the line.

[0013] Preferably, when constructing the virtual model of the substation line in the substation line virtual model construction module of the virtual repair simulation subunit in step S31, the accuracy of the model is adjusted according to the complexity of the fault. For general faults, a simplified line model is adopted, while for complex faults, a comprehensive virtual model is constructed.

[0014] In summary, compared with the prior art, the present invention provides a substation line fault diagnosis and processing system and method using artificial intelligence, having the following beneficial effects: 1. By installing the distributed sensor nodes and quantum sensors in the monitoring module at the positions to be monitored on the substation line, the present invention can collect the temperature, current, and voltage data of the line in real time, which helps to comprehensively understand the operating state of the line, provides a rich information basis for subsequent fault diagnosis, and thus more accurately locates possible faults. Moreover, through the coupling of the micro energy harvesting device integrated inside the distributed sensor nodes and quantum sensors with the substation electromagnetic field, self-sufficiency of energy is achieved, enabling the sensors to operate without external continuous power supply, reducing the complexity and cost of wiring, while improving the independence and stability of the system, ensuring that the data acquisition work can be carried out continuously and stably; 2. The environmental adaptability subunit added inside the monitoring module of the present invention can adjust the sensors according to the environmental conditions of the substation. The substation environment is complex and changeable, and different environmental factors such as temperature, humidity, and electromagnetic field may affect the performance of the sensors. The presence of the environmental adaptability subunit enables the sensors to maintain good working conditions under various environmental conditions, thereby ensuring the accuracy of the collected data, reducing data errors caused by environmental factors. At the same time, the graph neural network and physical knowledge embedding model set in the artificial intelligence algorithm unit in the diagnosis module can perform complex operations on the collected data through the relationship between its nodes and edges, effectively mining the characteristic information in the data. The thermophysical knowledge embedding model integrates the physical knowledge of the substation line into the diagnosis process. The combination of the two greatly improves the accuracy of identifying the fault type and location, and can quickly and accurately determine the type and occurrence location of the line fault; 3. Through the built-in online learning function of the diagnosis module, the present invention can dynamically update the model parameters according to new fault cases, enabling the diagnosis module to continuously learn and adapt to these new situations, continuously improving the accuracy of diagnosis, maintaining the system's response ability to various fault situations, and the virtual repair simulation subunit inside the processing module simulates the repair process in a virtual environment before performing actual operations, and evaluates the feasibility and the generated impact of the repair plan according to the simulation results. This simulation process can detect possible problems in the repair plan in advance, avoid unnecessary losses and risks during the actual repair process, and improve the reliability of fault handling. Brief Description of the Drawings

[0015] Figure 1 is a schematic structural diagram of the present invention.

[0016] Figure 2 is a schematic structural diagram of the present invention.

[0017] Figure 3 is a schematic structural diagram of the present invention.

[0018] Figure 4 is a schematic structural diagram of the present invention. Detailed Embodiment

[0019] 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. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0020] Please refer to Figure 1 , the present invention provides a technical solution, a substation line fault diagnosis and processing system using artificial intelligence, including: A monitoring module, a diagnosis module, and a processing module. The monitoring module includes distributed sensor nodes and quantum sensors. The monitoring module is installed at the position to be monitored on the substation line. The monitoring module is used to collect data on the temperature, current, and voltage of the line in real time. An environmental adaptability sub-unit is additionally provided inside the monitoring module. Miniature energy harvesting devices are integrated inside both the distributed sensor nodes and the quantum sensors. The miniature energy harvesting devices are coupled with the substation electromagnetic field. The distributed sensor nodes include temperature sensors and Hall effect current sensors. Inductive probes are installed outside both the temperature sensors and the Hall effect current sensors. The inductive probes are distributed in an array. A protective shell with electromagnetic shielding function is provided outside the quantum sensor. The outer layer of the protective shell with electromagnetic shielding function is a metal shielding layer. The middle layer of the protective shell with electromagnetic shielding function is an insulating buffer layer. The inner layer of the protective shell with electromagnetic shielding function is a heat-conducting layer. The environmental adaptability sub-unit includes a miniature sensor and an intelligent control chip. The intelligent control chip is connected to the distributed sensor nodes and the quantum sensors; The diagnosis module is internally provided with an artificial intelligence algorithm unit. The diagnosis module is used to analyze the data collected by the monitoring module and identify the fault type and location. The artificial intelligence algorithm unit is provided with a graph neural network and a physical knowledge embedding model. An online learning function is built in the diagnosis module, which can dynamically update the model parameters according to new fault cases. The nodes in the graph neural network represent the various components and monitoring points in the substation line. The edges in the graph neural network represent the electrical connection relationships and physical associations between these components and monitoring points. A weight matrix is provided for each layer of the graph neural network. The structure of the physical knowledge embedding model in the artificial intelligence algorithm unit is a hierarchical structure. The bottom layer of the physical knowledge embedding model is the basic physical parameter layer, which contains the basic physical characteristic parameters of the substation line. The middle layer of the physical knowledge embedding model is the physical process simulation layer; The processing module is internally equipped with an automated control unit. The processing module is used to generate processing instructions based on the diagnostic results and perform fault isolation and repair operations. The processing module is internally equipped with a virtual repair simulation sub-unit. The virtual repair simulation sub-unit is used to simulate the repair process in a virtual environment before performing the actual fault isolation and repair operations, evaluate the feasibility and the generated impact of the repair plan according to the simulation results. The virtual repair simulation sub-unit includes a substation line virtual model construction module, a fault injection module, and a repair operation simulation module. The substation line virtual model construction module is used to construct a virtual substation line model according to the actual layout of the substation, the line connection relationship, and the device parameter information. The fault injection module is used to inject corresponding faults into the virtual model according to the fault type and location obtained by the diagnostic module. The repair operation simulation module is used to perform repair operation simulation in the virtual model according to the processing instructions generated by the processing module.

[0021] The present invention discloses a method for diagnosing and processing substation line faults using artificial intelligence, including the following steps; S1 Monitoring module data acquisition step: First, the temperature sensor and the Hall effect current sensor in the distributed sensor nodes monitor the substation line. At the same time, the micro sensors in the environmental adaptability sub-unit collect the environmental parameters inside the substation. Subsequently, the intelligent control chip adjusts the working states of the distributed sensor nodes and the quantum sensors according to the collected environmental parameters. The specific implementation steps of the above process are as follows; First, perform sensor installation and preparation: Determine the key nodes of the substation line according to the historical fault data, electrical topology, and heat exchange characteristics of the substation line. Install the temperature sensor and the Hall effect current sensor in the distributed sensor nodes at these key nodes. The sensing probes of the temperature sensor and the Hall effect current sensor are distributed in an array to ensure that the relevant physical quantities of the line can be comprehensively and accurately sensed. Install the quantum sensor at the position to be monitored on the substation line. The quantum sensor is covered with a protective shell with electromagnetic shielding function. The outer layer of the protective shell is a metal shielding layer, the middle layer is an insulating buffer layer, and the inner layer is a heat-conducting layer to ensure the normal operation of the quantum sensor and ensure that the micro sensors in the environmental adaptability sub-unit are in a workable state. The micro sensors can collect the environmental parameters inside the substation; Initial monitoring and environmental parameter collection: The temperature sensor and the Hall effect current sensor in the distributed sensor nodes start monitoring the substation line at the same time. The temperature sensor starts measuring the temperature of the line, and the Hall effect current sensor starts measuring the current in the line. The micro sensors in the environmental adaptability sub-unit start collecting the environmental parameters inside the substation at the same time, such as parameters such as the temperature, humidity, air pressure, dust content, and corrosive gas concentration inside the substation; Environmental parameter analysis and sensor status adjustment: The micro-sensor transmits the collected environmental parameters to the intelligent control chip, which analyzes the received environmental parameters and analyzes the possible impacts of the environmental parameters on the operation of the sensors. For example, if the environmental temperature is too high, it may affect the accuracy of the temperature sensor; if the humidity is high, it may affect the insulation performance of electrical equipment, and further affect the measurement accuracy of the Hall effect current sensor, etc. According to the analysis results of the environmental parameters, the intelligent control chip adjusts the operating status of the distributed sensor nodes (including temperature sensors and Hall effect current sensors) and quantum sensors. For the temperature sensor, if the environmental temperature is too high, the intelligent control chip may reduce the sensitivity range of its measurement to avoid the measured value exceeding the sensor range or increase the measurement accuracy range for more accurate measurement in the current environment. For the Hall effect current sensor, if the environmental humidity is high, the intelligent control chip may adjust the internal amplification circuit parameters to compensate for the impact of humidity on the measurement. For the quantum sensor, if there is strong electromagnetic interference in the substation (judged by the electromagnetic-related indicators in the environmental parameters), the intelligent control chip may adjust some calibration parameters inside the quantum sensor to ensure the accuracy of its measurement; Finally, the collected data is transmitted to the diagnosis module. The temperature sensor and the Hall effect current sensor are installed at the key nodes of the substation line, and the key nodes are determined according to the historical fault data, electrical topology structure, and heat exchange characteristics of the line; S2 Diagnostic module fault analysis steps: First, after receiving the data collected in step S1, the diagnostic module performs preliminary preprocessing on the received data, and the preprocessing methods include; Data reception and verification: The diagnostic module receives the data transmitted from the monitoring module, and these data include the line temperature and current data collected by the distributed sensor nodes (temperature sensors and Hall effect current sensors), the data collected by the quantum sensor, and the environmental parameter data collected by the environmental adaptability subunit micro-sensor, etc. Check whether the received data is complete, and judge by means of the data format, length, and preset check code, etc. For example, check whether the temperature sensor data is within a reasonable numerical range and data format. If the data is missing or the format is incorrect, mark this part of the data as abnormal data, check the time stamp of the data to ensure the correct timing of the data. If it is found that the time order of the data is chaotic, it may affect subsequent analysis, and the data needs to be re-sorted or marked as abnormal; Data cleaning: Identify outliers in the data and process data points that deviate significantly from the normal range. For example, if the temperature value collected by the temperature sensor exceeds the upper limit of the normal operating temperature of the substation line by a large margin (such as exceeding a certain multiple of the historical maximum temperature), it is determined as an outlier. After the diagnostic module receives the data collected in step S1, it performs preliminary preprocessing on the received data. Due to the complex substation environment, the collected data may contain noise. For noise in data such as current and temperature, filtering techniques can be used for processing. For example, for current data, if there is high-frequency noise, a low-pass filter can be used to retain the low-frequency effective current signal and remove the high-frequency noise components; S21 Graph neural network analysis: Subsequently, the nodes and edges in the graph neural network perform operations based on the received data, and the data is weighted according to the actual situation of the line through the weight matrix of each layer to analyze the feature information in the data and preliminarily determine the features related to the fault area and fault type; S22 Physical knowledge embedding model analysis: The data processed in step S21 enters the physical knowledge embedding model and is matched at the bottom layer of basic physical parameters. Subsequently, at the intermediate layer of physical process simulation, the physical process in the line is simulated based on the basic physical parameters to further analyze the fault type and location; S23 Comprehensive diagnosis and model update: The diagnostic module synthesizes the analysis results of the graph neural network and the physical knowledge embedding model to identify the fault type and location. When a new fault case appears, the online learning function of the diagnostic module will dynamically update the model parameters according to the new case; S3 Fault handling steps of the processing module: S31 Virtual repair simulation: After receiving the data in step S2, the virtual repair simulation subunit in the processing module is activated, enabling the fault injection module to inject the corresponding fault into the virtual model according to the fault type and location obtained by the diagnostic module. Subsequently, the repair operation simulation module follows the processing instructions generated by the processing module according to the diagnostic results and performs repair operation simulation in the virtual model. When the substation line virtual model construction module of the virtual repair simulation subunit constructs the virtual model, the accuracy of the model is adjusted according to the complexity of the fault. For general faults, a simplified line model is used, while for complex faults, a comprehensive virtual model is constructed. The specific process of the above method is as follows; Virtual repair simulation subunit startup and virtual model preparation: After the processing module receives the analysis results (such as fault type and location information) from the diagnostic module, it starts the internal virtual repair simulation subunit. If the virtual substation line model has not been constructed before, the virtual substation line model construction module constructs a virtual substation line model based on the actual layout of the substation, line connection relationships, and equipment parameter information, and determines the model accuracy according to the complexity of the fault. For general faults, a simplified line model is adopted, which includes the main line structure and key equipment, and some minor details are ignored; for complex faults, a comprehensive virtual model is constructed, and the model reflects various parameters, equipment connections, and operating status information of the substation line as detailed as possible; Fault injection module operation: The fault injection module obtains fault type and location information from the diagnostic module. For example, if the diagnostic module identifies a short - circuit fault in a current transformer at a specific node in the substation line, the fault injection module obtains detailed information such as the fault type being "short - circuit" and the fault location being "the current transformer at a specific node". According to the fault type, the fault injection module determines the parameters required to simulate the fault in the virtual model. For short - circuit faults, parameters such as the short - circuit resistance value (which may be a value close to zero) and the line nodes involved need to be determined; for open - circuit faults, relevant parameters such as the disconnected line connection points need to be determined. The fault injection module injects the corresponding fault in the virtual substation line model according to the determined parameters. For example, in the virtual model, change the circuit connection mode of the current transformer at a specific node to simulate a short - circuit fault, or disconnect the connection of a specific line to simulate an open - circuit fault, etc.; Operation of the repair operation simulation module: The repair operation simulation module obtains the processing instructions generated according to the diagnostic results from the processing module. The processing instructions may include operation contents such as cutting off the power supply of a specific circuit, replacing a faulty device, adjusting the parameters of the protection device, etc. According to the processing instructions, the repair operation simulation module determines the operation steps and tools required to perform the repair operation in the virtual model (simulation tools in the virtual environment). For example, if the processing instruction is to replace a faulty device, the repair operation simulation module needs to determine how to simulate the operation steps such as disassembling the device, installing the new device, and connecting in the virtual model. The repair operation simulation module performs the repair operation simulation in the virtual model according to the determined operation steps. For example, in the virtual model, first cut off the power connection of the faulty device, then disassemble the faulty device, install the new device and reconnect the circuit, and adjust the parameters of the relevant protection device, etc., to simulate the entire repair process. During the simulation process, various data during the operation are recorded, such as the impact of the operation on other circuits and devices (such as whether it causes voltage fluctuations, whether it affects the normal operation of other devices, etc.), the time required for the operation, etc. The repair operation simulation module sorts out various data during the simulated repair operation to prepare for the subsequent evaluation of the feasibility and impact of the repair plan. For example, summarize the data such as the voltage fluctuation range caused by the operation and the change in the device operation state for comparative analysis with the preset safety standards and normal operation requirements; S32 Actual fault isolation and repair: When the result of the virtual repair simulation shows that the repair plan is feasible, the automatic control unit in the processing module generates processing instructions according to the diagnostic results and performs fault isolation and repair operations.

[0022] In this solution, the distributed sensor nodes and quantum sensors in the monitoring module are installed at the positions to be monitored on the substation line, which can collect the temperature, current, and voltage data of the line in real time, helping to comprehensively understand the operation state of the line, providing a rich information basis for subsequent fault diagnosis, thus more accurately locating possible faults. And through the internal integration of the micro energy harvesting device in the distributed sensor nodes and quantum sensors with the substation electromagnetic field coupling, the self-sufficiency of energy is achieved, enabling the sensors to operate without external continuous power supply, reducing the complexity and cost of wiring, while improving the independence and stability of the system, ensuring that the data acquisition work can be carried out continuously and stably.

[0023] Through the environmental adaptability subunit added inside the monitoring module, this solution can adjust the sensors according to the environmental conditions of the substation. The environment of the substation is complex and changeable, and different environmental factors such as temperature, humidity, electromagnetic field, etc. may affect the performance of the sensors. The existence of the environmental adaptability subunit enables the sensors to maintain a good working state under various environmental conditions, thereby ensuring the accuracy of the collected data, reducing data errors caused by environmental factors. At the same time, the graph neural network and physical knowledge embedding model set in the artificial intelligence algorithm unit in the diagnosis module can perform complex operations on the collected data through the relationship between its nodes and edges, effectively mining the feature information in the data. The thermal physics knowledge embedding model integrates the physical knowledge of the substation lines into the diagnosis process. The combination of the two greatly improves the accuracy of identifying the fault type and location, and can quickly and accurately determine the type and occurrence location of the line fault.

[0024] This solution also has an online learning function built into the diagnosis module that can dynamically update the model parameters according to new fault cases, enabling the diagnosis module to continuously learn and adapt to these new situations, continuously improving the accuracy of diagnosis, and maintaining the system's response ability to various fault situations. And the virtual repair simulation subunit inside the processing module simulates the repair process in a virtual environment before performing actual operations, and evaluates the feasibility and impact of the repair plan according to the simulation results. This simulation process can discover potential problems with the repair plan in advance, avoid unnecessary losses and risks during the actual repair process, and improve the reliability of fault handling.

[0025] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not expressly listed, or elements inherent to such process, method, article or device.

[0026] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A substation line fault diagnosis and processing system using artificial intelligence, characterized in that, Including: A monitoring module, a diagnostic module, and a processing module. The monitoring module includes distributed sensor nodes and quantum sensors. The monitoring module is installed at the position to be monitored on the substation line. The monitoring module is used to collect data on the temperature, current, and voltage of the line in real time. An environmental adaptability subunit is additionally provided inside the monitoring module. Miniature energy harvesting devices are integrated inside both the distributed sensor nodes and the quantum sensors. The miniature energy harvesting devices are coupled with the substation electromagnetic field; The diagnostic module is internally provided with an artificial intelligence algorithm unit. The diagnostic module is used to analyze the data collected by the monitoring module, identify the fault type and location. The artificial intelligence algorithm unit is provided with a graph neural network and a physical knowledge embedding model. An online learning function is built into the diagnostic module, which can dynamically update the model parameters according to new fault cases; The processing module is internally equipped with an automatic control unit. The processing module is used to generate a processing instruction according to the diagnostic result and perform fault isolation and repair operations. A virtual repair simulation subunit is installed inside the processing module. The virtual repair simulation subunit is used to simulate the repair process in a virtual environment before performing the actual fault isolation and repair operations, and evaluate the feasibility and the generated impact of the repair plan according to the simulation result.

2. The substation line fault diagnosis and processing system using artificial intelligence according to claim 1, characterized in that: The distributed sensor nodes include temperature sensors and Hall effect current sensors. Inductive probes are installed outside both the temperature sensors and the Hall effect current sensors. The inductive probes are distributed in an array.

3. The substation line fault diagnosis and processing system using artificial intelligence according to claim 1, characterized in that: The outside of the quantum sensor is covered with a protective shell with electromagnetic shielding function. The outer layer of the protective shell with electromagnetic shielding function is a metal shielding layer. The middle of the protective shell with electromagnetic shielding function is an insulating buffer layer. The inner layer of the protective shell with electromagnetic shielding function is a heat-conducting layer.

4. A substation line fault diagnosis and processing system using artificial intelligence according to claim 1, characterized in that: The environmental adaptability subunit includes a miniature sensor and an intelligent control chip. The intelligent control chip is connected to the distributed sensor nodes and the quantum sensors.

5. The substation line fault diagnosis and processing system using artificial intelligence according to claim 1, characterized in that: The nodes in the graph neural network represent each component and monitoring point in the substation line. The edges in the graph neural network represent the electrical connection relationships and physical associations between these components and monitoring points. Each layer of the graph neural network is provided with a weight matrix.

6. The substation line fault diagnosis and processing system using artificial intelligence according to claim 1, characterized in that: The structure of the physical knowledge embedding model in the artificial intelligence algorithm unit is a hierarchical structure. The bottom layer of the physical knowledge embedding model is the basic physical parameter layer, which contains the basic physical characteristic parameters of the substation line. The middle layer of the physical knowledge embedding model is the physical process simulation layer.

7. A substation line fault diagnosis and processing system using artificial intelligence according to claim 1, characterized in that: The virtual repair simulation subunit includes a virtual substation line model construction module, a fault injection module, and a repair operation simulation module. The virtual substation line model construction module is used to construct a virtual substation line model according to the actual layout of the substation, the line connection relationship, and the equipment parameter information. The fault injection module is used to inject corresponding faults into the virtual model according to the fault type and location obtained by the diagnostic module. The repair operation simulation module is used to perform a repair operation simulation in the virtual model according to the processing instruction generated by the processing module.

8. A substation line fault diagnosis and processing method using artificial intelligence, based on the substation line fault diagnosis and processing system using artificial intelligence according to any one of claims 1-7, characterized in that: Including the following steps: Data acquisition steps of the S1 monitoring module: First, the temperature sensor and Hall effect current sensor in the distributed sensor nodes monitor the substation line, and at the same time, the micro sensors in the environmental adaptability sub-unit collect the environmental parameters inside the substation. Subsequently, the intelligent control chip adjusts the working states of the distributed sensor nodes and quantum sensors according to the collected environmental parameters, and finally transmits the collected data to the diagnostic module; Fault analysis steps of the S2 diagnostic module: First, after receiving the data collected in step S1, the diagnostic module performs preliminary preprocessing on the received data; S21 Graph neural network analysis: Subsequently, the nodes and edges in the graph neural network perform operations according to the received data, and the weight matrix of each layer performs weighted processing on the data according to the actual situation of the line, analyzes the characteristic information in the data, and preliminarily judges the characteristics related to the fault area and fault type; S22 Physical knowledge embedding model analysis: The data processed in step S21 enters the physical knowledge embedding model and is matched at the bottom layer of basic physical parameters layer. Subsequently, at the middle layer of physical process simulation layer, the physical process in the line is simulated according to the basic physical parameters to further analyze the fault type and location; S23 Comprehensive diagnosis and model update: The diagnostic module comprehensively analyzes the results of the graph neural network and the physical knowledge embedding model to identify the fault type and location. When new fault cases occur, the online learning function of the diagnostic module will dynamically update the model parameters according to the new cases; Fault handling steps of the S3 processing module: S31 Virtual repair simulation: After receiving the data in step S2, the virtual repair simulation sub-unit in the processing module is started, so that the fault injection module injects corresponding faults into the virtual model according to the fault type and location obtained by the diagnostic module. Subsequently, the repair operation simulation module follows the processing instructions generated by the processing module according to the diagnostic results and performs repair operation simulation in the virtual model; S32 Actual fault isolation and repair: When the result of the virtual repair simulation shows that the repair plan is feasible, the automatic control unit in the processing module generates processing instructions according to the diagnostic results and performs fault isolation and repair operations.

9. A method for diagnosing and processing substation line faults using artificial intelligence according to claim 8, characterized in that: The temperature sensor and Hall effect current sensor are installed at the key nodes of the substation line, and the key nodes are determined according to the historical fault data, electrical topology structure and heat exchange characteristics of the line.

10. A method for diagnosing and processing substation line faults using artificial intelligence according to claim 8, characterized in that: When constructing the virtual model of the substation line in the virtual repair simulation sub-unit in step S31, the accuracy of the model is adjusted according to the complexity of the fault. For general faults, a simplified line model is adopted, while for complex faults, a comprehensive virtual model is constructed.

Images