Intelligent centralized control and protection integrated system for high-low voltage switch cabinet

Abstract

The invention discloses an intelligent centralized control and protection integrated system for a high-low voltage switch cabinet, relates to the technical field of electric power automation, and solves the problems that an existing centralized control system is lagged in data processing, incompatible in communication protocol and difficult in fault positioning. Running parameters and equipment state parameters of each high-low voltage switch cabinet are monitored in real time through a distributed data acquisition module; different communication protocols are automatically identified and adapted through the communication transmission module; real-time data analysis, fault early warning and state evaluation are carried out through a deep analysis module, and an early warning mechanism is triggered when a potential fault is early warned; the fault diagnosis module is used for quickly positioning a fault source and generating a maintenance suggestion, and the remote interaction module is used for providing a graphical user interface, displaying the real-time state of the system and allowing a user to remotely monitor and operate; according to the invention, the performance, compatibility and usability of the system are obviously improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electric power automation, and more specifically to an intelligent centralized control and protection integrated system for high and low voltage switch cabinets. Background Art

[0002] With the continuous improvement of the intelligence level of power systems, the operating status monitoring and fault diagnosis of power equipment have become particularly important. Modern power grids are facing growing power demand and more complex grid structures, which put higher requirements on the reliability of power equipment. As one of the key equipment in the power system, the stable operation of high and low voltage switchgear directly affects the safety and reliability of power supply. Therefore, it has become an inevitable trend to develop an intelligent high and low voltage switchgear centralized control and protection integrated system to achieve more efficient and intelligent management and maintenance of power equipment. Behind this trend is the deep integration of information technology and power technology, as well as the need to improve the automation level of the power system.

[0003] The integrated system of centralized control and protection of high and low voltage switchgear aims to improve the reliability and efficiency of the power system by integrating monitoring, protection, measurement and communication functions. Such systems usually use advanced sensor technology and data acquisition methods to monitor the status of power equipment in real time and predict potential faults through data analysis. For example, the document "Intelligent High Voltage Switchgear Online Monitoring System" monitors the changes in the internal environment and electrical parameters of the switchgear in real time by integrating temperature sensors, current transformers and other devices. In addition, the system also has the ability to communicate with the superior dispatching center or other automation equipment to achieve remote control and data sharing. Although the application of these technologies has significantly improved the management level of the power system, some problems that need to be solved in actual applications have been exposed. For example, when the system faces massive data, especially during peak hours, data processing may lag. When the system needs to process data from multiple switchgears at the same time, its computing power and storage capacity may not be sufficient to support efficient data analysis, thus affecting the timeliness of fault warning. In addition, the equipment provided by different manufacturers may use different communication protocols, which brings difficulties to system integration. For example, IEC 61850 and Modbus are two common industrial communication protocols, but there are large differences between them, which makes seamless communication between devices complicated; this not only increases the complexity of system integration, but also may cause consistency and integrity problems in data transmission; at the same time, due to the presence of multiple devices and protocols in the system, once a fault occurs, it becomes very difficult to locate the specific cause of the problem. For example, when the monitoring system reports an abnormality, technicians need to check each device and its communication link one by one, which greatly prolongs the time for fault diagnosis. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention discloses an integrated intelligent centralized control and protection system for high and low voltage switch cabinets to solve the problems in the background technology.

[0005] The present invention adopts the following technical solutions:

[0006] An integrated intelligent centralized control and protection system for high and low voltage switch cabinets, comprising:

[0007] A distributed data acquisition module is used to monitor the operating parameters and equipment status parameters of each high and low voltage switch cabinet in real time through a sensor network deployed in each switch cabinet, including voltage, current, temperature, switch status and smoke, and perform preliminary data processing through an embedded microcontroller, including data verification, filtering and formatting. The processed data is stored through a data cache unit, which stores the preliminary processed data through a first-in-first-out mechanism and uses an LZ77 compression algorithm to reduce the amount of data;

[0008] The communication transmission module is used to automatically identify and adapt the different communication protocols used by the connected high and low voltage switchgear devices through the multi-protocol conversion framework, convert the communication protocols of different devices into a unified internal protocol, and provide standardized data access services;

[0009] A deep analysis module is used to receive and process the data from the distributed data acquisition module, perform real-time data analysis, fault warning and status evaluation, and generate a status analysis report and fault prediction results; the deep analysis module triggers the early warning mechanism when a potential fault is warned, and sends fault information to the fault diagnosis module;

[0010] A fault diagnosis module, used to perform fault location, cause analysis and maintenance suggestion generation based on the fault information output by the in-depth analysis module; the fault diagnosis module includes a fault location submodule, a cause analysis submodule and a maintenance suggestion generation submodule;

[0011] The remote interaction module is used to provide a graphical user interface through the remote monitoring interface, display the real-time status of the system, and allow users to issue instructions or query information, including querying historical data, viewing reports, submitting feedback, remote switch control and remote parameter setting operations.

[0012] As a further technical solution of the present invention, the embedded microcontroller identifies abnormal data through threshold monitoring during data processing. If the detected data exceeds the set threshold, it is marked through interrupt processing and uploaded to the deep analysis module with priority.

[0013] As a further technical solution of the present invention, the real-time monitoring method and processing method of the embedded microcontroller are as follows: first, the analog signal output by the sensor is continuously read through a real-time data acquisition method, and the received analog signal is converted into a digital signal through an analog-to-digital converter; based on the converted digital signal, the integrity and accuracy of the data are ensured through a real-time data verification mechanism CRC, and the digital signal is dynamically filtered through an adaptive filter; based on the processed data, the embedded microcontroller continuously compares the current data with a preset normal operating parameter range through a preset threshold monitoring algorithm; if it is detected that the data exceeds the set threshold, the embedded microcontroller marks the abnormal data through an interrupt processing mechanism, and places the abnormal data at the front of the queue through a priority queue mechanism, and sends the abnormal data to the deep analysis module through the real-time communication protocol MQTT; during the abnormal data processing, the non-abnormal data is compressed through the LZ77 compression algorithm, and the abnormal data and non-abnormal data are sent to the data cache unit through a synchronous backup mechanism.

[0014] As a further technical solution of the present invention, the working principle of the communication transmission module is as follows: based on the data frame received from the high and low voltage switch cabinet equipment, the communication transmission module first analyzes the characteristics of the data frame through the device fingerprint, including the frame header, frame tail, data field and check code, and identifies the type of communication protocol used by the device; based on the identified protocol type, the various fields in the data frame are dynamically parsed through the multi-protocol conversion framework, and key information is extracted, including command words, data length, and check code; the multi-protocol conversion framework maps the parsed data field to a unified internal protocol format through a protocol mapping table to generate a new data packet; then, the communication transmission module encapsulates the new data packet into a unified data format through a data encapsulation mechanism, including adding header information and tail information; after encapsulation, the communication transmission module processes the encapsulated data packet through a multi-protocol stack and selects a transmission channel for transmission; during the transmission process, the communication transmission module ensures the reliability of path transmission through a redundant path transmission method; the communication transmission module also supports the upper-level management system to remotely monitor, control and protect the high and low voltage switch cabinet equipment through a standardized data access service interface.

[0015] As a further technical solution of the present invention, the working method of the remote monitoring, control and protection operation of the communication transmission module is as follows: based on the control instruction data packet issued by the remote interaction module, the communication transmission module first parses the header information of the data packet to extract the identifier of the target device, the control instruction type and related parameters; based on the extracted target device identifier, the communication protocol type used by the target device is identified through the device fingerprint, and the control instruction is converted into a specific format suitable for the target device through the protocol mapping table; then, the communication transmission module selects the communication protocol through the multi-protocol conversion framework and encapsulates the instruction into a data frame format; during the encapsulation process, a frame header, a frame footer and a checksum are added to the data frame. information to ensure the integrity and correctness of the data frame; after encapsulation is completed, the data frame is sent to the target device through the real-time communication protocol TCP / IP; after sending the control instruction, the communication transmission module waits for the confirmation reply of the target device through the sending status feedback mechanism; if no confirmation reply is received within the preset time, the control instruction is resent until the confirmation is received or the maximum number of retries is reached; once the confirmation reply of the target device is received, the module parses the confirmation reply data frame to verify whether the control instruction is successfully executed; if the control instruction is successfully executed, a success status report is generated; if the control instruction fails to execute, a failure report is generated and an error message is provided; after the report is generated, the execution status report is returned to the control system through the standardized interface RESTful API.

[0016] As a further technical solution of the present invention, the multi-protocol conversion framework includes a protocol feature extraction layer, a field parsing layer, a semantic understanding layer, a mapping rule matching layer, a data conversion layer, a verification encryption layer and a data output layer; the working steps of the multi-protocol conversion framework include:

[0017] Step 601: extracting protocol features of a data packet through the protocol feature extraction layer; the protocol feature extraction layer identifies and separates key features in a data frame through regular expression matching, including a frame header, a frame tail, a command word, a data length, and a checksum field;

[0018] Step 602: According to the extracted protocol features, the field parsing layer parses each field in the data packet; the field parsing layer extracts the command word, data length and check code information in the data frame through an adaptive protocol parsing algorithm, and stores the information as structured data;

[0019] Step 603: dynamically adjusting the parsing strategy according to the context of the data frame through the semantic understanding layer; the semantic understanding layer identifies different states in the data frame through a state machine and adjusts the parsing logic according to these states;

[0020] Step 604: according to the parsed key information and structured data, the mapping rule matching layer searches for the corresponding mapping rule in the protocol mapping table; the mapping rule matching layer searches for the corresponding rule in the protocol mapping table through the hash table;

[0021] Step 605: Based on the mapping rules and structured data found by the mapping rule matching layer, the mapped fields are arranged according to the format requirements of the internal protocol through the data conversion layer, and necessary header information, including protocol identifier and data length, is added to generate a new data packet;

[0022] Step 606: Verify and encrypt the new data packet through the verification encryption layer;

[0023] Step 607: Output the converted new data packet to an upper management system or other application layer services through the data output layer.

[0024] As a further technical solution of the present invention, the working method of the adaptive protocol parsing algorithm is:

[0025] Step 701: Load the corresponding protocol specification from the protocol library according to the protocol type identifier provided by the protocol feature extraction layer;

[0026] Step 702: Create a parsing context environment through a parsing context manager to store temporary data and status information during the parsing process, including the current parsing position and field counter, and set the initial value of these status information;

[0027] Step 703: Generate a field parsing list containing detailed information of all fields in the data packet, including field name, data type, length, and offset, according to the loaded protocol specification;

[0028] Step 704: traverse each field in the field resolution list; for each field in the list, perform the following operations:

[0029] 1) Extract the original data of the field from the data packet according to the offset and length of the field;

[0030] 2) According to the data type of the field, the extracted raw data is converted or decoded to obtain the actual value of the field;

[0031] 3) Store the actual value of the field in the specified location in the parsing context;

[0032] Step 705: When encountering a special field in the process of traversing the field parsing list, execute corresponding special processing logic according to the rules in the protocol specification; the special field includes a nested structure and a variable length field, and the special processing logic includes recursive parsing and dynamic calculation of field length;

[0033] Step 706: After parsing all fields, the checksum or check code of the data packet is calculated by the verifier according to the check rules in the protocol specification, and the checksum or check code is compared with the check value provided in the data packet. If the check fails, the error information is recorded and the error handling logic is triggered;

[0034] Step 707: Update the status information in the parsing context environment, including the field counter and the parsing completion flag, through the parsing context manager to reflect the parsing status of the current data packet;

[0035] Step 708: End the parsing process and integrate the data structure and information in the parsing context environment into a parsing result object, which includes the actual values ​​of all fields in the data packet and the verification result information.

[0036] As a further technical solution of the present invention, the working method of the deep analysis module is: receiving the initially processed data transmitted by the distributed data acquisition module through a real-time data receiving unit, and integrating the multi-source data from different sensors through a data fusion algorithm to eliminate redundant information and form a unified data view; based on the integrated data, the deep analysis module processes the data in real time through a stream processing framework to identify patterns and trends in the data; the stream processing framework performs pattern recognition and classification on the data through a support vector machine to determine whether there are abnormalities or potential faults; if abnormal data or potential fault patterns are detected; the deep analysis module triggers an early warning signal through an early warning mechanism and sends the fault information to a fault diagnosis module.

[0037] As a further technical solution of the present invention, the fault location submodule includes a data preprocessing unit, a feature extraction unit and a positioning algorithm unit; the data preprocessing unit removes noise and standardizes the format of the input data through data cleaning and preprocessing; the feature extraction unit extracts key features from the input data through a convolutional neural network; based on the extracted feature data, the positioning algorithm unit locates the fault source through a random forest; the cause analysis submodule includes a causal reasoning unit, a pattern recognition unit and a comprehensive analysis unit; the causal reasoning unit realizes the causal relationship inference between the fault phenomenon and the potential cause through a Bayesian network; the pattern recognition unit identifies the fault mode from the historical fault data through a support vector machine; the comprehensive analysis unit conducts an in-depth analysis of the cause of the fault by comprehensively considering multiple factors through a multi-factor analysis method; based on the fault identification result and the analysis result; the maintenance suggestion generation submodule matches the corresponding diagnosis rules according to the results of fault mode recognition through a rule-based expert system, and uses the fault cases and solutions in the knowledge base to provide a reference basis.

[0038] As a further technical solution of the present invention, the positioning algorithm unit locates the fault source through random forests in the following manner:

[0039] S1, based on the fault feature data set obtained from the deep analysis module;

[0040] S2. Based on the preprocessed fault feature data, the features related to the fault are extracted by a feature extraction method, including voltage fluctuation, current anomaly and temperature change index, and a set of fault state feature vectors are generated;

[0041] S3. Based on the fault state feature vector, an integrated learning model is constructed by a random forest algorithm. The integrated learning model extracts multiple training sample sets from the original data by a self-service sampling method, and trains a decision tree for each sample set. Each tree divides the data according to the feature importance, and the final classification result is determined by a majority voting mechanism.

[0042] S4, during the training process, node splitting is performed by random feature subset selection;

[0043] S5. In the process of building the decision tree, starting from the root node, the data set is divided into two left and right child nodes by recursive splitting according to the selected feature subset and splitting criteria, wherein the splitting criteria include Gini impurity and information gain; until the stopping condition is met, including that the number of samples in the node is less than the threshold or all samples belong to the same category;

[0044] S6. After all decision trees are constructed, a voting mechanism or weighted average method is used to combine the prediction results of all trees to obtain the final fault source location;

[0045] S7. During the fault location process, the random forest algorithm updates the model through incremental learning to adapt to the dynamic changes of fault modes.

[0046] The positive beneficial effects of the present invention are:

[0047] 1. The embedded microcontroller in the distributed data acquisition module performs preliminary data processing, and uses the data cache unit to store data through a first-in-first-out mechanism. At the same time, the LZ77 compression algorithm is used to reduce the amount of data. The present invention significantly improves data processing efficiency and reduces the demand for data storage capacity. This effectively solves the problems of data processing lag and insufficient storage capacity during peak periods. At the same time, abnormal data is marked through interrupt processing and uploaded to the deep analysis module in priority, ensuring the real-time and high efficiency of data processing, thereby improving the timeliness of fault warning.

[0048] 2. Through the multi-protocol conversion framework in the communication transmission module, the present invention can automatically identify and adapt the different communication protocols used by the connected high and low voltage switch cabinet equipment, convert the communication protocols of different devices into a unified internal protocol, and provide standardized data access services to control each high and low voltage switch cabinet equipment; this not only simplifies the complexity of system integration, but also ensures the consistency and integrity of data transmission, improves the compatibility and scalability of the system, and reduces the cost of system integration and maintenance.

[0049] 3. Through the deep analysis module and fault diagnosis module, the system can analyze data in real time, conduct fault warning and status evaluation, and send fault information to the fault diagnosis module when a potential fault is warned. The fault diagnosis module further locates the fault, analyzes the cause, and generates maintenance suggestions, which greatly shortens the time for fault diagnosis. At the same time, through the remote monitoring, control, and protection operations of the communication transmission module, online fault handling is performed, further improving the efficiency of fault handling; through the remote interaction module of the graphical user interface, users can easily view the real-time status of the system, query historical data, view reports, submit feedback, and perform remote switch control and parameter setting operations, enhancing the system's availability and user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a schematic diagram of the overall architecture of an intelligent centralized control and protection integrated system for high and low voltage switch cabinets of the present invention;

[0051] Figure 2 This is a flow chart of the positioning method of the positioning algorithm unit of the present invention. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] In the present invention, if Figure 1As shown: An integrated system of intelligent centralized control and protection of high and low voltage switch cabinets includes: a distributed data acquisition module, which is used to monitor the operating parameters and equipment status parameters of each high and low voltage switch cabinet in real time through a sensor network deployed in each switch cabinet, including voltage, current, temperature, switch status and smoke, and perform preliminary data processing through an embedded microcontroller, including data verification, filtering and formatting processing. The processed data is stored through a data cache unit, and the data cache unit stores the preliminary processed data through a first-in-first-out mechanism, and uses the LZ77 compression algorithm to reduce the data volume; during the data processing process, the embedded microcontroller identifies abnormal data through threshold monitoring. If the detected data exceeds the set threshold, it is marked through interrupt processing and uploaded to the deep analysis module with priority. The real-time monitoring method and processing method of the embedded microcontroller are as follows: first, the analog signal output by the sensor is continuously read through the real-time data acquisition method, and the received analog signal is converted into a digital signal through an analog-to-digital converter; based on the converted digital signal, the integrity and accuracy of the data are ensured through the real-time data verification mechanism CRC, and the digital signal is dynamically filtered through an adaptive filter; based on the processed data, the embedded microcontroller continuously compares the current data with the preset normal operating parameter range through a preset threshold monitoring algorithm; if the data is detected to exceed the set threshold, the embedded microcontroller marks the abnormal data through the interrupt processing mechanism, and puts the abnormal data at the front of the queue through the priority queue mechanism, and sends the abnormal data to the deep analysis module through the real-time communication protocol MQTT; in the abnormal data processing process, the non-abnormal data is compressed and processed through the LZ77 compression algorithm, and the abnormal data and non-abnormal data are sent to the data cache unit through the synchronous backup mechanism. In the specific implementation, the distributed data acquisition module uses the real-time data acquisition method, and the embedded microcontroller continuously reads the analog signal output by the sensor, and converts the received analog signal into a digital signal through an analog-to-digital converter (ADC). This process ensures the digitization of the original signal, which is convenient for subsequent processing. Secondly, the integrity and accuracy of the data are ensured through a real-time data verification mechanism (such as CRC verification), and the digital signal is dynamically filtered through an adaptive filter. The adaptive filter can dynamically adjust the filtering parameters according to the characteristics of the signal, remove noise and retain useful information. Thirdly, based on the processed data, the embedded microcontroller continuously compares the current data with the preset normal operating parameter range through a preset threshold monitoring algorithm. This process identifies abnormal data by comparing the current data with the preset threshold. Finally, if the data is detected to exceed the set threshold, the embedded microcontroller marks the abnormal data through an interrupt processing mechanism and puts the abnormal data at the front of the queue through a priority queue mechanism. The abnormal data is sent to the deep analysis module through a real-time communication protocol (such as MQTT).At the same time, the non-abnormal data is compressed by the LZ77 compression algorithm, and the abnormal data and non-abnormal data are sent to the data cache unit through the synchronous backup mechanism. This process ensures the priority processing of abnormal data and the effective storage of non-abnormal data.

[0054] In practical applications, the distributed data acquisition module monitors the operating parameters and equipment status parameters of each high and low voltage switch cabinet in real time through the sensor network deployed in each switch cabinet, including voltage, current, temperature, switch status and smoke. The module performs preliminary data processing through the embedded microcontroller, including data verification, filtering and formatting, and the processed data is stored through the data cache unit. The data cache unit stores the preliminarily processed data through the first-in-first-out (FIFO) mechanism and uses the LZ77 compression algorithm to reduce the amount of data. In actual work, the distributed data acquisition module can realize real-time monitoring of the operating status of high and low voltage switch cabinets, detect abnormal situations in time, and ensure the timely upload of key data through the priority processing mechanism. This method not only improves the efficiency of data processing, but also ensures the accuracy and integrity of the data. Compared with the existing data acquisition system, this module ensures the priority processing of abnormal data through the real-time monitoring and interrupt processing mechanism of the embedded microcontroller, and improves the response speed of the system. The storage space of non-abnormal data is reduced by the LZ77 compression algorithm, and the efficient storage and fast access of data are ensured by the FIFO mechanism, which improves the storage efficiency.

[0055] The communication transmission module is used to automatically identify and adapt the different communication protocols used by the connected high and low voltage switchgear devices through the multi-protocol conversion framework, convert the communication protocols of different devices into a unified internal protocol, and provide standardized data access services; in the specific implementation, the logical operation process of the communication transmission module is as follows:

[0056] First, the communication transmission module analyzes the characteristics of the data frame through device fingerprint technology to identify the type of communication protocol used by the device. This process involves analyzing the characteristics of the frame header, frame tail, data field, check code, etc. of the data frame, and comparing them with the feature library of the known protocol to determine the communication protocol to which the data frame belongs. Next, the communication transmission module dynamically parses each field in the data frame through an adaptive protocol parsing algorithm and extracts key information such as command words, data length, check code, etc. This process dynamically adjusts the parsing rules according to the identified protocol type to ensure that all necessary field information can be correctly extracted. Then, the communication transmission module maps the parsed data field to a unified internal protocol format through a protocol mapping table to generate a new data packet. This process involves converting the parsed data field into a data packet in a unified format according to the mapping table to ensure data consistency and processability. Finally, the communication transmission module encapsulates the converted data packet into a unified data format through a data encapsulation mechanism, and selects a suitable transmission channel for transmission through multi-protocol stack technology. This process involves adding necessary header information and tail information to the data packet, and selecting the best transmission path through multi-protocol stack technology to ensure the integrity and correctness of the data packet during transmission.

[0057] In the prior art, the communication transmission module usually needs to be manually configured or relies on a specific protocol conversion device, which leads to high integration complexity and prone to data consistency problems. For example, the traditional method may require writing a special conversion code for each protocol, which increases the complexity of the system. The present invention realizes the function of automatic identification and adaptation of different communication protocols through device fingerprint technology and adaptive protocol parsing algorithm, and simplifies the system integration process. In addition, the data conversion and encapsulation process in the prior art is not flexible enough and difficult to adapt to diverse needs. The present invention realizes the unification of data format and the reliability of transmission process through protocol mapping table and data encapsulation mechanism, and improves the flexibility and stability of the system. In a specific implementation, the working principle of the communication transmission module is as follows: based on the data frame received from the high and low voltage switch cabinet equipment, the communication transmission module first analyzes the characteristics of the data frame through the device fingerprint, including the frame header, frame tail, data field and check code, and identifies the type of communication protocol used by the device; based on the identified protocol type, the various fields in the data frame are dynamically parsed through the multi-protocol conversion framework, and key information, including command words, data length, and check code, are extracted; the multi-protocol conversion framework maps the parsed data field to a unified internal protocol format through a protocol mapping table to generate a new data packet; then, the communication transmission module encapsulates the new data packet into a unified data format through a data encapsulation mechanism, including adding header information and tail information; after encapsulation, the communication transmission module processes the encapsulated data packet through a multi-protocol stack and selects a transmission channel for transmission; during the transmission process, the communication transmission module ensures the reliability of path transmission through a redundant path transmission method; the communication transmission module also supports the upper management system to remotely monitor, control and protect the high and low voltage switch cabinet equipment through a standardized data access service interface;

[0058] The working method of the remote monitoring, control and protection operation of the communication transmission module is as follows: based on the control instruction data packet issued by the remote interaction module, the communication transmission module first parses the header information of the data packet to extract the identifier of the target device, the control instruction type and related parameters; based on the extracted target device identifier, the communication protocol type used by the target device is identified through the device fingerprint, and the control instruction is converted into a specific format suitable for the target device through the protocol mapping table; then, the communication transmission module selects the communication protocol through the multi-protocol conversion framework and encapsulates the instruction into a data frame format; during the encapsulation process, a frame header, a frame footer and check information are added to the data frame to ensure that the data frame integrity and correctness; after encapsulation, the data frame is sent to the target device through the real-time communication protocol TCP / IP; after sending the control instruction, the communication transmission module waits for the confirmation reply of the target device through the sending status feedback mechanism; if no confirmation reply is received within the preset time, the control instruction is resent until the confirmation is received or the maximum number of retries is reached; once the confirmation reply of the target device is received, the module parses the confirmation reply data frame to verify whether the control instruction is successfully executed; if the control instruction is successfully executed, a success status report is generated; if the control instruction fails to execute, a failure report is generated and an error message is provided; after the report is generated, the execution status report is returned to the control system through the standardized interface RESTful API;

[0059] The multi-protocol conversion framework includes a protocol feature extraction layer, a field parsing layer, a semantic understanding layer, a mapping rule matching layer, a data conversion layer, a verification encryption layer and a data output layer; the working steps of the multi-protocol conversion framework include:

[0060] Step 601: extracting protocol features of a data packet through the protocol feature extraction layer; the protocol feature extraction layer identifies and separates key features in a data frame through regular expression matching, including a frame header, a frame tail, a command word, a data length, and a checksum field;

[0061] Step 602: According to the extracted protocol features, the field parsing layer parses each field in the data packet; the field parsing layer extracts the command word, data length and check code information in the data frame through an adaptive protocol parsing algorithm, and stores the information as structured data;

[0062] Step 603: dynamically adjusting the parsing strategy according to the context of the data frame through the semantic understanding layer; the semantic understanding layer identifies different states in the data frame through a state machine and adjusts the parsing logic according to these states;

[0063] Step 604: according to the parsed key information and structured data, the mapping rule matching layer searches for the corresponding mapping rule in the protocol mapping table; the mapping rule matching layer searches for the corresponding rule in the protocol mapping table through the hash table;

[0064] Step 605: Based on the mapping rules and structured data found by the mapping rule matching layer, the mapped fields are arranged according to the format requirements of the internal protocol through the data conversion layer, and necessary header information, including protocol identifier and data length, is added to generate a new data packet;

[0065] Step 606: Verify and encrypt the new data packet through the verification encryption layer;

[0066] Step 607: Output the converted new data packet to an upper management system or other application layer services through the data output layer.

[0067] In practical applications, the multi-protocol conversion framework can realize automatic identification and adaptation of different communication protocols, convert the communication protocols of different devices into a unified internal protocol, and provide standardized data access services. This function solves the integration complexity and data consistency problems caused by different devices using different communication protocols in the existing technology. In practical applications, the multi-protocol conversion framework can significantly improve the system's integration efficiency and data processing capabilities, ensure seamless communication between different devices, and provide a consistent data format for subsequent data processing and fault diagnosis. This approach not only simplifies the system's integration process, but also improves the system's stability and reliability.

[0068] In the prior art, the communication transmission module usually needs to be manually configured or relies on a specific protocol conversion device, resulting in high integration complexity and prone to data consistency problems. For example, the traditional method may require writing special conversion codes for each protocol, which increases the complexity of the system. The present invention realizes the functions of automatic identification and adaptation of different communication protocols through regular expression matching technology, adaptive protocol parsing algorithm, state machine technology and hash table technology, and simplifies the integration process of the system. In addition, the data conversion and encapsulation process in the prior art is not flexible enough and difficult to adapt to diverse needs. The present invention arranges the mapped fields according to the format requirements of the internal protocol, and ensures the integrity and security of the data through verification encryption technology, thereby improving the flexibility and stability of the system.

[0069] Furthermore, in practical applications, the adaptive protocol parsing algorithm can achieve efficient parsing of data packets of different communication protocols and convert the parsed data into a unified internal format. This function solves the data consistency problem caused by different devices using different communication protocols in the prior art. In practical applications, the adaptive protocol parsing algorithm can significantly improve the parsing efficiency and data processing capabilities of the system, ensure seamless communication between different devices, and provide a consistent data format for subsequent data processing and fault diagnosis. This approach not only simplifies the system integration process, but also improves the stability and reliability of the system.

[0070] In the specific implementation, step 701 loads the corresponding protocol specification from the protocol library according to the protocol type identifier provided by the protocol feature extraction layer. This process selects the appropriate protocol definition file from the protocol library through the protocol type identifier (for example, IEC 61850, Modbus, etc.). The protocol specification contains detailed information such as field name, data type, length, offset, etc., ensuring that the parsing algorithm can parse for a specific protocol type. The protocol library is a predefined database that contains standard definitions of various communication protocols. It identifies the protocol type through regular expression matching technology and loads the corresponding protocol specification from the protocol library. Step 702 creates a parsing context environment through a parsing context manager to store temporary data and status information during the parsing process, including the current parsing position and field counter, and sets the initial value of these status information. The parsing context manager ensures the consistency and integrity of the status information during the parsing process. The parsing context environment contains information such as the current parsing position and the number of parsed fields. By initializing these status information, the continuity and accuracy of the parsing process are ensured. Step 703 generates a field parsing list containing detailed information of all fields in the data packet according to the loaded protocol specification. This process generates a detailed field parsing list through the protocol specification, ensuring the clarity and accuracy of the field information. The field parsing list is a list containing information such as field name, data type, length, offset, etc. Through this information, the parsing algorithm can accurately know how to extract and process the data of each field. Step 704 traverses each field in the field parsing list, extracts the original data of the field from the data packet according to the offset and length of the field, and performs type conversion or decoding processing according to the data type of the field to obtain the actual value of the field. This process ensures the correct extraction and type conversion of field data. Field extraction locates the position of the field in the data packet by offset and extracts data according to length. Type conversion performs corresponding conversion processing according to the data type of the field (such as integer, floating point, etc.) to ensure data consistency and accuracy. Step 705 When encountering special fields (such as nested structures and variable-length fields) during the traversal of the field parsing list, the corresponding special processing logic is executed according to the rules in the protocol specification, such as recursive parsing and dynamic calculation of field length. This processing logic ensures the correct parsing of complex fields. Special field processing handles nested structures through recursive parsing and handles variable-length fields through dynamic calculation of field length, ensuring that the parsing algorithm can handle complex field structures. Step 706 After parsing all fields, according to the verification rules in the protocol specification, the checksum or check code of the data packet is calculated by the verifier, and the checksum or check code is compared with the check value provided in the data packet to ensure the integrity of the data. This process ensures the integrity and correctness of the data packet through the verification rules.The verification rule is usually part of the protocol specification. The verification value is calculated through a verification algorithm (such as CRC verification) and compared with the verification value in the data packet to ensure the integrity and correctness of the data. Step 707 updates the status information in the parsing context environment through the parsing context manager, including the field counter and the parsing completion flag, to reflect the parsing status of the current data packet. This process ensures the correct recording of the parsing status. The parsing context update ensures the continuity and consistency of the parsing state by updating the field counter and the parsing completion flag. Step 708 ends the parsing process and integrates the data structure and information in the parsing context environment into a parsing result object, which contains the actual values ​​of all fields in the data packet and the verification result information. This process ensures the consistency and integrity of the parsing results. The parsing result integration integrates the actual values ​​and verification result information of all fields into a structured object for subsequent processing and storage.

[0071] In the prior art, protocol parsing usually requires manual configuration or relies on a specific parser, which makes the parsing process complicated and prone to data consistency problems. For example, the traditional method may require writing special parsing code for each protocol, which increases the complexity of the system. The present invention realizes the functions of automatic identification and adaptation of different communication protocols through protocol specification loading, parsing context management, field parsing list generation and special field processing logic, and simplifies the system integration process. In addition, the data parsing and verification process in the prior art is not flexible enough and difficult to adapt to diverse needs. The present invention ensures the accuracy and integrity of the data through field parsing lists and verification rules, and improves the flexibility and stability of the system.

[0072] A deep analysis module is used to receive and process the data of the distributed data acquisition module, perform real-time data analysis, fault warning and status evaluation, and generate a status analysis report and fault prediction results; the deep analysis module triggers the warning mechanism when a potential fault is warned, and sends fault information to the fault diagnosis module; in a specific implementation, the working method of the deep analysis module is: receiving the preliminary processed data transmitted by the distributed data acquisition module through a real-time data receiving unit, and integrating multi-source data from different sensors through a data fusion algorithm to eliminate redundant information and form a unified data view; based on the integrated data, the deep analysis module processes the data in real time through a stream processing framework to identify patterns and trends in the data; the stream processing framework performs pattern recognition and classification on the data through a support vector machine to determine whether there is an abnormality or potential fault; if abnormal data or potential fault patterns are detected; the deep analysis module triggers a warning signal through the warning mechanism, and sends the fault information to the fault diagnosis module.

[0073] In actual applications, the deep analysis module can realize real-time monitoring and abnormal detection of the operating status of high and low voltage switchgear, and generate status analysis reports and fault prediction results. This function solves the problem of untimely fault warning caused by insufficient data processing capabilities in the existing technology. In actual applications, the deep analysis module can significantly improve the system's real-time monitoring capabilities and the accuracy of fault warnings, ensuring that the system can detect potential faults in a timely manner and take preventive measures, thereby improving the reliability and stability of the system.

[0074] In specific implementation, the logical operation process of the deep analysis module is as follows:

[0075] Real-time data reception: The deep analysis module receives the preliminarily processed data transmitted by the distributed data acquisition module through the real-time data receiving unit. This process ensures the real-time and reliability of data transmission.

[0076] Data fusion: Data fusion algorithms are used to integrate multi-source data from different sensors, eliminate redundant information, and form a unified data view. This process ensures the consistency and integrity of the data.

[0077] Real-time processing and pattern recognition: Based on the integrated data, the deep analysis module processes the data in real time through the stream processing framework to identify patterns and trends in the data. This process ensures the real-time analysis capability of the data through real-time data processing technology.

[0078] Pattern recognition and classification: The stream processing framework uses support vector machines to perform pattern recognition and classification on the data to determine whether there are anomalies or potential failures. This process uses the learning ability of support vector machines to identify abnormal data or potential failure patterns.

[0079] Early warning mechanism triggering: If abnormal data or potential failure modes are detected, the deep analysis module triggers an early warning signal through the early warning mechanism and sends the failure information to the fault diagnosis module. This process ensures that the early warning signal can be triggered in time when an abnormality is detected, and the relevant information is passed to the fault diagnosis module.

[0080] In the prior art, data processing usually requires manual configuration or relies on specific data processing tools, which makes the data processing process complicated and prone to data consistency problems. For example, the traditional method may require writing a special data processing script for each data source, which increases the complexity of the system. The present invention realizes the function of automatically identifying and processing multi-source data through a real-time data receiving unit, a data fusion algorithm, a stream processing framework and a support vector machine, and simplifies the system integration process.

[0081] A fault diagnosis module is used to perform fault location, cause analysis and maintenance suggestion generation based on the fault information output by the deep analysis module; the fault diagnosis module includes a fault location submodule, a cause analysis submodule and a maintenance suggestion generation submodule; the fault location submodule includes a data preprocessing unit, a feature extraction unit and a positioning algorithm unit; the data preprocessing unit removes noise and standardizes the format of the input data through data cleaning and preprocessing; the feature extraction unit extracts key features from the input data through a convolutional neural network; based on the extracted feature data, the positioning algorithm unit locates the fault source through a random forest; the cause analysis submodule includes a causal reasoning unit, a pattern recognition unit and a comprehensive analysis unit; the causal reasoning unit realizes the causal relationship inference between the fault phenomenon and the potential cause through a Bayesian network; the pattern recognition unit identifies the fault mode from the historical fault data through a support vector machine; the comprehensive analysis unit conducts an in-depth analysis of the fault cause by comprehensively considering multiple factors through a multi-factor analysis method; based on the fault identification result and the analysis result; the maintenance suggestion generation submodule matches the corresponding diagnosis rules according to the result of fault mode identification through a rule-based expert system, and provides a reference basis using the fault cases and solutions in the knowledge base. The positioning algorithm unit locates the fault source by using random forest as follows:

[0082] S1, based on the fault feature data set obtained from the deep analysis module;

[0083] S2. Based on the preprocessed fault feature data, the features related to the fault are extracted by a feature extraction method, including voltage fluctuation, current anomaly and temperature change index, and a set of fault state feature vectors are generated;

[0084] S3. Based on the fault state feature vector, an integrated learning model is constructed by a random forest algorithm. The integrated learning model extracts multiple training sample sets from the original data by a self-service sampling method, and trains a decision tree for each sample set. Each tree divides the data according to the feature importance, and the final classification result is determined by a majority voting mechanism.

[0085] S4, during the training process, node splitting is performed by random feature subset selection;

[0086] S5. In the process of building the decision tree, starting from the root node, the data set is divided into two left and right child nodes by recursive splitting according to the selected feature subset and splitting criteria, wherein the splitting criteria include Gini impurity and information gain; until the stopping condition is met, including that the number of samples in the node is less than the threshold or all samples belong to the same category;

[0087] S6. After all decision trees are constructed, a voting mechanism or weighted average method is used to combine the prediction results of all trees to obtain the final fault source location;

[0088] S7. During the fault location process, the random forest algorithm updates the model through incremental learning to adapt to the dynamic changes of the fault mode. Figure 2 As shown in Figure 1, the data set of the positioning algorithm unit comes from the deep analysis module. After obtaining the data, the next step is to preprocess these fault feature data to ensure the quality of the data and make it more suitable for subsequent feature extraction. After the preprocessing is completed, the system will extract the features related to the fault, including key indicators such as voltage fluctuation, current anomaly and temperature change, to form the fault state feature vectors, which are the basis for building the random forest model. Subsequently, the system will build a random forest model based on these feature vectors. The model uses the self-service sampling method to extract multiple training sample sets from the original data to provide a data basis for the training of each tree. During the training process of each decision tree, the system will select a random feature subset, and then divide the data set into two left and right child nodes by recursive splitting according to the selected feature subset and splitting criteria (such as Gini impurity and information gain). This process will continue until the stopping condition is met, such as the number of samples in the node is less than the preset threshold or all samples belong to the same category. When all decision trees are built and trained, the system will synthesize the prediction results of all trees. This step is usually carried out by voting mechanism or weighted average to determine the final fault source location result. In practical applications, the fault mode may change dynamically over time, so the random forest model needs to be continuously updated through incremental learning to adapt to these changes and ensure the accuracy and robustness of the model. Finally, when the model update is completed, the entire fault source location process is completed, providing accurate location information for fault diagnosis and repair.

[0089] The remote interaction module is used to provide a graphical user interface through the remote monitoring interface, display the real-time status of the system, and allow users to issue instructions or query information, including querying historical data, viewing reports, submitting feedback, remote switch control and remote parameter setting operations.

[0090] During system operation, the high and low voltage switchgear intelligent centralized control and protection integrated system always maintains real-time data collection, processing and analysis to ensure continuous monitoring of system status. When an abnormality or fault is detected, the system immediately starts the early warning and fault diagnosis process, quickly locates the fault point, and provides detailed fault analysis reports and maintenance suggestions. Users can view system status, historical data, fault reports, etc. in real time through the remote interaction module, and perform remote operations or adjust system parameters as needed.

[0091] Although the specific embodiments of the present invention are described above, it should be understood by those skilled in the art that these specific embodiments are only illustrative, and those skilled in the art may omit, replace, and change the details of the above methods and systems in various ways without departing from the principles and essence of the present invention. For example, merging the above method steps so as to perform substantially the same functions in substantially the same manner to achieve substantially the same results is within the scope of the present invention. Therefore, the scope of the present invention is limited only by the appended claims.

Claims

1. An intelligent centralized control and protection integrated system for high and low voltage switch cabinets, characterized in that: include: A distributed data acquisition module is used to monitor the operating parameters and equipment status parameters of each high and low voltage switch cabinet in real time through a sensor network deployed in each switch cabinet, including voltage, current, temperature, switch status and smoke, and perform preliminary data processing through an embedded microcontroller, including data verification, filtering and formatting. The processed data is stored through a data cache unit, which stores the preliminary processed data through a first-in-first-out mechanism and uses an LZ77 compression algorithm to reduce the amount of data; The communication transmission module is used to automatically identify and adapt the different communication protocols used by the connected high and low voltage switchgear devices through the multi-protocol conversion framework, convert the communication protocols of different devices into a unified internal protocol, and provide standardized data access services; A deep analysis module is used to receive and process the data from the distributed data acquisition module, perform real-time data analysis, fault warning and status evaluation, and generate a status analysis report and fault prediction results; the deep analysis module triggers the early warning mechanism when a potential fault is warned, and sends fault information to the fault diagnosis module; A fault diagnosis module, used to perform fault location, cause analysis and maintenance suggestion generation based on the fault information output by the in-depth analysis module; the fault diagnosis module includes a fault location submodule, a cause analysis submodule and a maintenance suggestion generation submodule; The remote interaction module is used to provide a graphical user interface through the remote monitoring interface, display the real-time status of the system, and allow users to issue instructions or query information, including querying historical data, viewing reports, submitting feedback, remote switch control and remote parameter setting operations.

2. According to claim 1, a high and low voltage switch cabinet intelligent centralized control and protection integrated system is characterized by: During the data processing, the embedded microcontroller identifies abnormal data through threshold monitoring. If the detected data exceeds the set threshold, it is marked through interrupt processing and uploaded to the deep analysis module with priority.

3. According to claim 1, a high and low voltage switch cabinet intelligent centralized control and protection integrated system is characterized by: The real-time monitoring method and processing method of the embedded microcontroller are as follows: first, the analog signal output by the sensor is continuously read through a real-time data acquisition method, and the received analog signal is converted into a digital signal through an analog-to-digital converter; based on the converted digital signal, the integrity and accuracy of the data are ensured through a real-time data verification mechanism CRC, and the digital signal is dynamically filtered through an adaptive filter; based on the processed data, the embedded microcontroller continuously compares the current data with the preset normal operating parameter range through a preset threshold monitoring algorithm; if it is detected that the data exceeds the set threshold, the embedded microcontroller marks the abnormal data through an interrupt processing mechanism, and places the abnormal data at the front of the queue through a priority queue mechanism, and sends the abnormal data to a deep analysis module through a real-time communication protocol MQTT; during the abnormal data processing, the non-abnormal data is compressed through the LZ77 compression algorithm, and the abnormal data and non-abnormal data are sent to the data cache unit through a synchronous backup mechanism.

4. According to claim 1, a high and low voltage switch cabinet intelligent centralized control and protection integrated system is characterized by: The working principle of the communication transmission module is as follows: based on the data frame received from the high and low voltage switch cabinet equipment, the communication transmission module first analyzes the characteristics of the data frame through the device fingerprint, including the frame header, frame tail, data field and check code, and identifies the type of communication protocol used by the device; based on the identified protocol type, the various fields in the data frame are dynamically parsed through the multi-protocol conversion framework, and key information, including command words, data length, and check code, are extracted; the multi-protocol conversion framework maps the parsed data field to a unified internal protocol format through a protocol mapping table to generate a new data packet; then, the communication transmission module encapsulates the new data packet into a unified data format through a data encapsulation mechanism, including adding header information and tail information; after encapsulation, the communication transmission module processes the encapsulated data packet through a multi-protocol stack and selects a transmission channel for transmission; during the transmission process, the communication transmission module ensures the reliability of path transmission through a redundant path transmission method; the communication transmission module also supports the upper management system to remotely monitor, control and protect the high and low voltage switch cabinet equipment through a standardized data access service interface.

5. The high and low voltage switch cabinet intelligent centralized control and protection integrated system according to claim 4 is characterized in that: The working method of the remote monitoring, control and protection operation of the communication transmission module is as follows: based on the control instruction data packet issued by the remote interaction module, the communication transmission module first parses the header information of the data packet to extract the identifier of the target device, the control instruction type and related parameters; based on the extracted target device identifier, the communication protocol type used by the target device is identified through the device fingerprint, and the control instruction is converted into a specific format suitable for the target device through the protocol mapping table; then, the communication transmission module selects the communication protocol through the multi-protocol conversion framework and encapsulates the instruction into a data frame format; during the encapsulation process, a frame header, a frame footer and check information are added to the data frame to ensure the integrity and correctness of the data frame; After the encapsulation is completed, the data frame is sent to the target device through the real-time communication protocol TCP / IP; after sending the control instruction, the communication transmission module waits for the confirmation reply of the target device through the sending status feedback mechanism; if no confirmation reply is received within the preset time, the control instruction is resent until the confirmation is received or the maximum number of retries is reached; Once a confirmation response is received from the target device, the module parses the confirmation response data frame to verify whether the control instruction is successfully executed; if the control instruction is successfully executed, a success status report is generated; if the control instruction fails to execute, a failure report is generated and error information is provided; after the report is generated, the execution status report is returned to the control system through the standardized interface RESTful API.

6. The high and low voltage switch cabinet intelligent centralized control and protection integrated system according to claim 1 is characterized in that: The multi-protocol conversion framework includes a protocol feature extraction layer, a field parsing layer, a semantic understanding layer, a mapping rule matching layer, a data conversion layer, a verification encryption layer and a data output layer; The working steps of the multi-protocol conversion framework include: Step 601: extracting protocol features of a data packet through the protocol feature extraction layer; the protocol feature extraction layer identifies and separates key features in a data frame through regular expression matching, including a frame header, a frame tail, a command word, a data length, and a checksum field; Step 602: According to the extracted protocol features, the field parsing layer parses each field in the data packet; the field parsing layer extracts the command word, data length and check code information in the data frame through an adaptive protocol parsing algorithm, and stores the information as structured data; Step 603: dynamically adjusting the parsing strategy according to the context of the data frame through the semantic understanding layer; the semantic understanding layer identifies different states in the data frame through a state machine and adjusts the parsing logic according to these states; Step 604: according to the parsed key information and structured data, the mapping rule matching layer searches for the corresponding mapping rule in the protocol mapping table; the mapping rule matching layer searches for the corresponding rule in the protocol mapping table through the hash table; Step 605: Based on the mapping rules and structured data found by the mapping rule matching layer, the mapped fields are arranged according to the format requirements of the internal protocol through the data conversion layer, and necessary header information, including protocol identifier and data length, is added to generate a new data packet; Step 606: Verify and encrypt the new data packet through the verification encryption layer; Step 607: Output the converted new data packet to an upper management system or other application layer services through the data output layer.

7. The high and low voltage switch cabinet intelligent centralized control and protection integrated system according to claim 6 is characterized in that: The working method of the adaptive protocol analysis algorithm is: Step 701: Load the corresponding protocol specification from the protocol library according to the protocol type identifier provided by the protocol feature extraction layer; Step 702: Create a parsing context environment through a parsing context manager to store temporary data and status information during the parsing process, including the current parsing position and field counter, and set the initial value of these status information; Step 703: Generate a field parsing list containing detailed information of all fields in the data packet, including field name, data type, length, and offset, according to the loaded protocol specification; Step 704: traverse each field in the field resolution list; for each field in the list, perform the following operations: 1) Extract the original data of the field from the data packet according to the offset and length of the field; 2) According to the data type of the field, the extracted raw data is converted or decoded to obtain the actual value of the field; 3) Store the actual value of the field in the specified location in the parsing context; Step 705: When encountering a special field in the process of traversing the field parsing list, execute corresponding special processing logic according to the rules in the protocol specification; the special field includes a nested structure and a variable length field, and the special processing logic includes recursive parsing and dynamic calculation of field length; Step 706: After parsing all fields, the checksum or check code of the data packet is calculated by the verifier according to the check rules in the protocol specification, and the checksum or check code is compared with the check value provided in the data packet. If the check fails, the error information is recorded and the error handling logic is triggered; Step 707: Update the status information in the parsing context environment, including the field counter and the parsing completion flag, through the parsing context manager to reflect the parsing status of the current data packet; Step 708: End the parsing process and integrate the data structure and information in the parsing context environment into a parsing result object, which includes the actual values ​​of all fields in the data packet and the verification result information.

8. The high and low voltage switch cabinet intelligent centralized control and protection integrated system according to claim 1 is characterized by: The working method of the in-depth analysis module is as follows: receiving the initially processed data transmitted by the distributed data acquisition module through the real-time data receiving unit, and integrating the multi-source data from different sensors through the data fusion algorithm, eliminating redundant information, and forming a unified data view; Based on the integrated data, the deep analysis module processes the data in real time through a stream processing framework to identify patterns and trends in the data; the stream processing framework performs pattern recognition and classification on the data through a support vector machine to determine whether there are abnormalities or potential faults; if abnormal data or potential fault patterns are detected; the deep analysis module triggers an early warning signal through an early warning mechanism and sends the fault information to the fault diagnosis module.

9. The high and low voltage switch cabinet intelligent centralized control and protection integrated system according to claim 1 is characterized by: The fault location submodule includes a data preprocessing unit, a feature extraction unit and a location algorithm unit; the data preprocessing unit removes noise and standardizes the format of the input data through data cleaning and preprocessing; the feature extraction unit extracts key features from the input data through a convolutional neural network; Based on the extracted feature data, the positioning algorithm unit locates the fault source through random forest; the cause analysis submodule includes a causal reasoning unit, a pattern recognition unit and a comprehensive analysis unit; the causal reasoning unit realizes the causal relationship inference between the fault phenomenon and the potential cause through the Bayesian network; the pattern recognition unit identifies the fault mode from the historical fault data through the support vector machine; the comprehensive analysis unit conducts in-depth analysis of the fault cause by comprehensively considering multiple factors through a multi-factor analysis method; Based on the fault identification results and analysis results; the maintenance suggestion generation submodule matches the corresponding diagnosis rules according to the results of fault pattern identification through a rule-based expert system, and uses the fault cases and solutions in the knowledge base to provide reference basis.

10. The high and low voltage switch cabinet intelligent centralized control and protection integrated system according to claim 9, characterized in that: The positioning algorithm unit locates the fault source by using random forest as follows: S1, based on the fault feature data set obtained from the deep analysis module; S2. Based on the preprocessed fault feature data, the features related to the fault are extracted by a feature extraction method, including voltage fluctuation, current anomaly and temperature change index, and a set of fault state feature vectors are generated; S3. Based on the fault state feature vector, an integrated learning model is constructed by a random forest algorithm. The integrated learning model extracts multiple training sample sets from the original data by a self-service sampling method, and trains a decision tree for each sample set. Each tree divides the data according to the feature importance, and the final classification result is determined by a majority voting mechanism. S4, during the training process, node splitting is performed by random feature subset selection; S5. In the process of building the decision tree, starting from the root node, the data set is divided into two left and right child nodes by recursive splitting according to the selected feature subset and splitting criteria, wherein the splitting criteria include Gini impurity and information gain; until the stopping condition is met, including that the number of samples in the node is less than the threshold or all samples belong to the same category; S6. After all decision trees are constructed, a voting mechanism or weighted average method is used to combine the prediction results of all trees to obtain the final fault source location; S7. During the fault location process, the random forest algorithm updates the model through incremental learning to adapt to the dynamic changes of fault modes.

Images