Relay protection control method based on multi-source data flow and related equipment

Abstract

The invention discloses a relay protection control method based on a multi-source data stream and related equipment, and belongs to the technical field of relay protection, electrical quantity data and switch state data are respectively collected, multiple key aspects of operation of a power system can be covered, and different types of data are preprocessed through a data transmission processing module, so that the accuracy of relay protection is improved. Meanwhile, the constructed hybrid transmission network performs shunt transmission according to data characteristics, guarantees the reliability of the data in the transmission process, extracts electrical quantity and switch state data feature vectors through an optimization fusion module, and fuses the electrical quantity and switch state data feature vectors into a comprehensive feature vector. The comprehensive feature vector can enable the model to more accurately judge the fault type, then based on a mode acquisition module, an improved support vector machine model is utilized, kernel function parameters and penalty factors are optimized through a particle swarm optimization algorithm, the classification accuracy and stability of the model are improved, the comprehensive feature vector serves as input, and the classification accuracy and stability of the model are improved. And the fault behavior mode of the data can be accurately output.

Description

Technical Field

[0001] The present invention belongs to the technical field of relay protection, and specifically relates to a relay protection control method based on multi-source data streams and related equipment. Background Art

[0002] With the continuous development of power systems and the advancement of intelligent processes, the requirements for relay protection control systems are also increasing. In modern power systems, the synchronous analysis of multiple data streams has become the key to achieving efficient and accurate relay protection. In traditional relay protection systems, they often only focus on a single type of data, such as electrical quantity data such as current and voltage. However, with the increasing complexity of power systems, it is difficult to rely solely on electrical quantity data to meet the comprehensive and accurate judgment of system faults. Discrete data such as switch status also contains rich system operation information. For example, the opening and closing state of the switch can directly reflect the operating status and fault location of the equipment; At the same time, the power system environment is complex, and different types of data have different characteristics and units. Electrical quantity data is usually continuous time series data, and its change law is crucial for analyzing the dynamic operation status of the system. Digital quantity data such as switch status is a discrete state signal, and its instantaneous change may indicate a major event in the system.

[0003] In the prior art, due to the complexity of the power system environment, different types of data have different characteristics and units. For example, electrical quantity data such as current and voltage are usually continuous time series data, while digital quantity data such as switch status are discrete state signals. Processing such highly diverse power data forms at the same time leads to a prolonged development cycle and it is difficult to ensure the stability and reliability of subsequent decision-making. Therefore, a relay protection control system based on multi-source data streams is proposed herein. Summary of the invention

[0004] The present invention provides a relay protection control method based on multi-source data streams and related equipment, which solves the problem of long development cycle and instability caused by simultaneously processing power data forms with great differences.

[0005] To achieve the above object, the present invention provides the following technical solutions: A relay protection control method based on multi-source data streams, comprising: Collect electrical quantity data and switch status data; Preprocessing electrical quantity data and switch status data, and building a hybrid transmission network based on the preprocessed electrical quantity data and switch status data; Extracting electrical quantity and switch state data feature vectors from the hybrid transmission network, and fusing the electrical quantity and switch state data feature vectors into a comprehensive feature vector; Based on the constructed support vector machine model, the comprehensive feature vector is used as the input of the improved support vector machine model, and the fault behavior pattern of the output data is obtained; Determine whether the fault behavior mode conforms to the pre-set fault behavior mode and whether a fault occurs in the power system.

[0006] Preferably, the process of collecting electrical quantity data and switch status data is: The electrical quantity data includes current and voltage data; The current data is collected based on the sensor, and the voltage data is collected using the voltage divider resistor network. The sampling frequency is set to ,in, , The highest frequency of the measured signal is used to obtain continuous electrical quantity time series data ,in, Indicates the type of electrical quantity. Indicates time; The switch state data adopts a photoelectric coupler combined with a digital input circuit, and a pulse signal is generated each time the state changes, and the time point and state value of the state change are recorded to obtain the switch state data.

[0007] Preferably, the construction process of the hybrid transmission network is: Based on the preprocessed electrical quantity data, the time series analysis algorithm is used to analyze its periodicity and trend changes, and the flow size in different time periods is predicted. Based on the switch status data, according to the equipment operation frequency and the number of historical state changes, its flow demand is estimated, and network bandwidth resources are allocated based on the prediction results. According to the real-time requirements, the electrical quantity data and key switch status data transmission are allocated to the frequency band or channel with interference less than the set threshold to obtain a hybrid transmission network.

[0008] Preferably, the steps of preprocessing the electrical quantity data and the switch state data are: The weighted moving average filter is used to smooth the electrical quantity data. The specific formula for the processing is: ; in is the filtered value, is the window size of the moving average, Represents the sequence number of discrete time points in a time series; The switch state data is processed by jitter and a time threshold is set. When the time interval between two adjacent state changes is less than the threshold, it is considered to be a false trigger and the original state value is maintained.

[0009] Preferably, the steps for constructing the improved support vector machine model are: Optimize the kernel function parameters and penalty factors of the basic support vector machine. Define the solution space and search space, obtain the particle position vector and particle velocity vector from the solution space, and the initial position and initial velocity of the particle are randomly generated in the search space; Define the particle's own historical optimal position and the group's historical optimal position to update the particle position vector and particle velocity vector; The particle updates its position vector based on the new velocity ; in, Indicates The particle position vector of the iteration; Each particle compares its current fitness value with its historical optimal fitness value. If the current fitness value is better, it updates its historical optimal position. At the same time, it compares the historical optimal fitness values ​​of all particles in the entire particle group to find the historical optimal position of the group. An iteration termination threshold is set, and the iteration is stopped when the iteration termination threshold is reached. The parameter combination corresponding to the historical optimal position of the group, namely the kernel function parameter and the penalty factor, is output as the optimization parameter of the support vector machine model. The support vector machine model is retrained using the optimized parameters to obtain an improved support vector machine model.

[0010] Preferably, the updating process of the particle velocity vector is:

[0011] in, is the number of iterations, Indicates that the inertia weight controls the degree to which the particle inherits the current velocity. and is the learning factor, which adjusts the step size of the particle approaching its own historical optimal position and the group's historical optimal position respectively. and are two random numbers.

[0012] Preferably, the step of acquiring the fault behavior mode is: The comprehensive feature vector is extracted and input into the trained improved support vector machine model. The improved support vector machine model determines which fault behavior mode the comprehensive feature vector sample belongs to through the internal optimized kernel function and penalty factor, and outputs the corresponding results, including normal operation, short circuit fault and equipment overload.

[0013] A relay protection control system based on multi-source data streams, comprising: Data acquisition module: used to collect electrical quantity data and switch status data; Data transmission processing module: used to pre-process electrical quantity data and switch status data, and build a hybrid transmission network based on the pre-processed electrical quantity data and switch status data; Optimization fusion module: used to extract electrical quantity and switch state data feature vectors from the hybrid transmission network, and fuse the electrical quantity and switch state data feature vectors into a comprehensive feature vector; Pattern acquisition module: used to take the comprehensive feature vector as the input of the improved support vector machine model based on the constructed support vector machine model, and output the fault behavior pattern of the data; Abnormal diagnosis module: used to determine whether the fault behavior mode conforms to the pre-set fault behavior mode and whether the power system has a fault.

[0014] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of a relay protection control method based on multi-source data streams are implemented.

[0015] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of a relay protection control method based on multi-source data streams are implemented.

[0016] Compared with the prior art, the present invention has the following beneficial effects: the present invention proposes a relay protection control method based on multi-source data streams, which collects electrical quantity data and switch status data respectively, can cover multiple key aspects of power system operation, and performs preprocessing operations on different types of data through the data transmission processing module, thereby improving data quality. At the same time, the constructed hybrid transmission network performs shunting transmission according to data characteristics, thereby ensuring the reliability of data during transmission. The electrical quantity and switch status data feature vectors are extracted through the optimization fusion module, and are fused into a comprehensive feature vector, which enables the model to more accurately determine the fault type. Then, based on the pattern acquisition module, the improved support vector machine model is used to optimize the kernel function parameters and penalty factors through the particle swarm optimization algorithm, thereby improving the classification accuracy and stability of the model. The comprehensive feature vector is used as input, and the fault behavior pattern of the data can be accurately output. The safe and stable operation of power system equipment is ensured. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 The present invention is a flow chart of a relay protection control method based on multi-source data streams.

[0018] Figure 2 This is a block diagram of a relay protection control system based on multi-source data streams of the present invention. DETAILED DESCRIPTION In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0019] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. 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.

[0020] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0021] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0022] like Figure 1 As shown, an embodiment of the present invention provides a relay protection control method based on multi-source data streams, including: S101 collects electrical quantity data and switch status data; S102 preprocesses the electrical quantity data and the switch status data, and builds a hybrid transmission network based on the preprocessed electrical quantity data and the switch status data; S103 extracts electrical quantity and switch state data feature vectors from the hybrid transmission network, and fuses the electrical quantity and switch state data feature vectors into a comprehensive feature vector; S104 uses the comprehensive feature vector as the input of the improved support vector machine model based on the constructed support vector machine model and outputs the fault behavior pattern of the data; S105 determines whether the fault behavior mode conforms to a preset fault behavior mode, and determines whether a fault occurs in the power system.

[0023] The process of collecting electrical quantity data and switch status data is as follows: Electrical quantity data include current and voltage data; The current data is collected based on the sensor, and the voltage data is collected using the voltage divider resistor network. The sampling frequency is set to ,in, , The highest frequency of the measured signal is used to obtain continuous electrical quantity time series data ,in, Indicates the type of electrical quantity. Indicates time; The switch state data uses a photocoupler combined with a digital input circuit. Each time the state changes, a pulse signal is generated, and the time point and state value of the state change are recorded to obtain the switch state data.

[0024] The construction process of the hybrid transmission network is as follows: Based on the preprocessed electrical quantity data, the time series analysis algorithm is used to analyze its periodicity and trend changes, and the flow size in different time periods is predicted. Based on the switch status data, according to the equipment operation frequency and the number of historical state changes, its flow demand is estimated, and network bandwidth resources are allocated based on the prediction results. According to the real-time requirements, the electrical quantity data and key switch status data transmission are allocated to the frequency band or channel with interference less than the set threshold to obtain a hybrid transmission network.

[0025] The steps for preprocessing electrical quantity data and switch status data are as follows: The weighted moving average filter is used to smooth the electrical quantity data. The specific formula for the processing is: ; in is the filtered value, is the window size of the moving average, Represents the sequence number of discrete time points in a time series; The switch state data is processed by jitter and a time threshold is set. When the time interval between two adjacent state changes is less than the threshold, it is considered to be a false trigger and the original state value is maintained.

[0026] By collecting electrical quantity data and switch status data separately, it can cover multiple key aspects of power system operation. For electrical quantity data, high-precision Hall sensors are used to collect current data, and voltage data is collected by using a voltage divider resistor network combined with a high-precision analog-to-digital converter to ensure the accuracy of continuous time series data. The dynamic changes of current and voltage in the power system can be monitored in real time, providing basic data support for analyzing the system operation status and fault detection. For discrete data such as switch status, optocouplers are used in combination with digital input circuits to accurately record the time point and state value of state changes, which helps to grasp the operation status and action of the equipment. When a fault occurs, it can quickly determine whether the operation related to the switch is correct. The data transmission processing module performs preprocessing operations on different types of data to improve data quality. It performs filtering processing on electrical quantity data to remove noise interference and make the data smoother, which is conducive to subsequent analysis; it performs de-jitter processing on discrete data to avoid false triggering and ensure the accuracy of switch status data. At the same time, a hybrid transmission network based on Ethernet and wireless private network is constructed, and shunting transmission is performed according to data characteristics. Ethernet ensures high-speed and stable transmission of continuous electrical quantity data, and wireless private network reduces the cost of discrete data transmission and improves flexibility. Data verification algorithms such as cyclic redundancy check are used to ensure data integrity and guarantee the reliability of data during transmission.

[0027] The fusion steps of the comprehensive feature vector are: For continuous electrical quantity data, the feature vector is extracted and expressed as , is the index of the electrical quantity data feature; The discrete switch state data forms a feature vector , is the index of the switch state data feature; The two types of data feature vectors are used as the input of the improved support vector machine model, expressed as .

[0028] Kernel function parameters for the basic support vector machine and penalty factor To optimize, Define a solution space and search space, and obtain the particle position vector from the solution space and the particle velocity vector , the initial position and initial velocity of the particle are randomly generated in the search space; Define a historical optimal position and the historical optimal position of the group Update particle position vector and the particle velocity vector ; Particle velocity vector The update process is: ; in, is the number of iterations, Indicates that the inertia weight controls the degree to which the particle inherits the current velocity. and is the learning factor, which adjusts the step size of the particle approaching its own historical optimal position and the group's historical optimal position respectively. and are two random numbers, The particle updates its position vector based on the new velocity ; in, Indicates The particle position vector of the iteration; Each particle compares its current fitness value with its historical optimal fitness value. If the current fitness value is better, it updates its historical optimal position. At the same time, compare the historical optimal fitness values ​​of all particles in the entire particle group to find the historical optimal position of the group ; An iteration termination threshold is set, and the iteration is stopped when the iteration termination threshold is reached. The parameter combination corresponding to the historical optimal position of the group, namely the kernel function parameter and the penalty factor, is output as the optimization parameters of the support vector machine model. The support vector machine model is retrained using these optimized parameters to obtain an improved support vector machine model.

[0029] The steps to obtain the fault behavior mode are: The comprehensive feature vector is extracted and input into the trained improved support vector machine model. The improved support vector machine model determines which fault behavior mode the comprehensive feature vector sample belongs to through the internal optimized kernel function and penalty factor, and outputs the corresponding results, including normal operation, short circuit fault and equipment overload.

[0030] Failure behavior modes include: The improved support vector machine model outputs the corresponding fault behavior mode, and the output results are "short circuit fault" and "overvoltage fault", as well as specific fault types and fault diagnosis information.

[0031] By inputting the comprehensive feature vector into the support vector machine model, this comprehensive feature vector contains various features extracted from the electrical quantity data and the switch state data, such as the mean, variance, harmonic content of the current, and the duration and change frequency of the switch state. In order to ensure that different features are numerically comparable, the input feature vector is standardized, and then the internal calculation of the support vector machine model is improved. The support vector machine maps the input feature vector to a high-dimensional space through a kernel function. For the input new feature vector and each sample feature vector in the training set, the distance or similarity measure between them is calculated through the kernel function, and then the optimization solution is performed. In the high-dimensional space, the support vector machine searches for an optimal classification hyperplane so that data with different fault behavior modes can be classified. The features of the new eigenvector are separated as accurately as possible. In the optimization process, the support vector machine is used to determine the position and direction of the hyperplane. By adjusting the kernel function parameters and the penalty factor, the classification hyperplane can have good generalization ability while ensuring the classification accuracy, that is, it can accurately classify new unknown data. After internal calculation, the support vector machine determines which fault behavior mode the new eigenvector belongs to according to its position in the high-dimensional space. Specifically, the category to which it belongs is determined by calculating the distance from the new eigenvector to the classification hyperplane and the relationship with the support vector. If the distance between the new eigenvector and the classification hyperplane corresponding to a certain fault behavior mode is the shortest and meets specific classification rules (such as being located on one side of the classification hyperplane), it is judged as a fault and relay protection is performed.

[0032] like Figure 2 The embodiment of the present invention further provides a relay protection control system based on multi-source data streams, comprising: Data acquisition module: used to collect electrical quantity data and switch status data; Data transmission processing module: used to pre-process electrical quantity data and switch status data, and build a hybrid transmission network based on the pre-processed electrical quantity data and switch status data; Optimization fusion module: used to extract electrical quantity and switch state data feature vectors from the hybrid transmission network, and fuse the electrical quantity and switch state data feature vectors into a comprehensive feature vector; Pattern acquisition module: used to take the comprehensive feature vector as the input of the improved support vector machine model based on the constructed support vector machine model, and output the fault behavior pattern of the data; Abnormal diagnosis module: used to determine whether the fault behavior mode conforms to the pre-set fault behavior mode and whether the power system has a fault.

[0033] The process of collecting electrical quantity data and switch status data is as follows: Electrical quantity data include current and voltage data; The current data is collected based on the sensor, and the voltage data is collected using the voltage divider resistor network. The sampling frequency is set to ,in, , The highest frequency of the measured signal is used to obtain continuous electrical quantity time series data ,in, Indicates the type of electrical quantity. Indicates time; The switch state data uses a photocoupler combined with a digital input circuit. Each time the state changes, a pulse signal is generated, and the time point and state value of the state change are recorded to obtain the switch state data.

[0034] The construction process of the hybrid transmission network is as follows: Based on the preprocessed electrical quantity data, the time series analysis algorithm is used to analyze its periodicity and trend changes, and the flow size in different time periods is predicted. Based on the switch status data, according to the equipment operation frequency and the number of historical state changes, its flow demand is estimated, and network bandwidth resources are allocated based on the prediction results. According to the real-time requirements, the electrical quantity data and key switch status data transmission are allocated to the frequency band or channel with interference less than the set threshold to obtain a hybrid transmission network.

[0035] The steps for preprocessing electrical quantity data and switch status data are as follows: The weighted moving average filter is used to smooth the electrical quantity data. The specific formula for the processing is: ; in is the filtered value, is the window size of the moving average, Represents the sequence number of discrete time points in a time series; The switch state data is processed by jitter and a time threshold is set. When the time interval between two adjacent state changes is less than the threshold, it is considered to be a false trigger and the original state value is maintained.

[0036] The steps for building the improved support vector machine model are: Optimize the kernel function parameters and penalty factors of the basic support vector machine. Define the solution space and search space, obtain the particle position vector and particle velocity vector from the solution space, and the initial position and initial velocity of the particle are randomly generated in the search space; Define the particle's own historical optimal position and the group's historical optimal position to update the particle position vector and particle velocity vector; The particle updates its position vector based on the new velocity ; in, Indicates The particle position vector of the iteration; Each particle compares its current fitness value with its historical optimal fitness value. If the current fitness value is better, it updates its historical optimal position. At the same time, it compares the historical optimal fitness values ​​of all particles in the entire particle group to find the historical optimal position of the group. An iteration termination threshold is set, and the iteration is stopped when the iteration termination threshold is reached. The parameter combination corresponding to the historical optimal position of the group, namely the kernel function parameter and the penalty factor, is output as the optimization parameter of the support vector machine model. The support vector machine model is retrained using the optimized parameters to obtain an improved support vector machine model.

[0037] The updating process of particle velocity vector is:

[0038] in, is the number of iterations, Indicates that the inertia weight controls the degree to which the particle inherits the current velocity. and is the learning factor, which adjusts the step size of the particle approaching its own historical optimal position and the group's historical optimal position respectively. and are two random numbers.

[0039] The steps to obtain the fault behavior mode are: The comprehensive feature vector is extracted and input into the trained improved support vector machine model. The improved support vector machine model determines which fault behavior mode the comprehensive feature vector sample belongs to through the internal optimized kernel function and penalty factor, and outputs the corresponding results, including normal operation, short circuit fault and equipment overload.

[0040] A terminal device is provided in one embodiment of the present invention. The terminal device of this embodiment includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps in the above-mentioned method embodiments are implemented. Alternatively, when the processor executes the computer program, the functions of the modules / units in the above-mentioned device embodiments are implemented.

[0041] The computer program may be divided into one or more modules / units, and the one or more modules / units are stored in the memory and executed by the processor to accomplish the present invention.

[0042] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0043] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0044] The memory may be used to store the computer programs and / or modules, and the processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory.

[0045] If the module / unit integrated in the terminal device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0046] Although the embodiments of the present invention are described above in conjunction with the accompanying drawings, the present invention is not limited to the above-mentioned specific embodiments and application fields, and the above-mentioned specific embodiments are only illustrative and instructive, rather than restrictive. Under the guidance of the specification, a person skilled in the art can also make many forms without departing from the scope of protection of the claims of the present invention, all of which belong to the scope of protection of the present invention.

Claims

1. A relay protection control method based on multi-source data stream, characterized in that: include: Collect electrical quantity data and switch status data; Preprocessing electrical quantity data and switch status data, and building a hybrid transmission network based on the preprocessed electrical quantity data and switch status data; Extracting electrical quantity and switch state data feature vectors from the hybrid transmission network, and fusing the electrical quantity and switch state data feature vectors into a comprehensive feature vector; Based on the constructed support vector machine model, the comprehensive feature vector is used as the input of the improved support vector machine model, and the fault behavior pattern of the output data is obtained; Determine whether the fault behavior mode conforms to the pre-set fault behavior mode and whether a fault occurs in the power system.

2. A relay protection control method based on multi-source data stream according to claim 1, characterized in that: The process of collecting electrical quantity data and switch status data is as follows: The electrical quantity data includes current and voltage data; The current data is collected based on the sensor, and the voltage data is collected using the voltage divider resistor network. The sampling frequency is set to ,in, , The highest frequency of the measured signal is used to obtain continuous electrical quantity time series data ,in, Indicates the type of electrical quantity. Indicates time; The switch state data adopts a photoelectric coupler combined with a digital input circuit, and each state change generates a pulse signal, and the time point and state value of the state change are recorded to obtain the switch state data.

3. A relay protection control method based on multi-source data stream according to claim 1, characterized in that: The construction process of the hybrid transmission network is as follows: Based on the preprocessed electrical quantity data, the time series analysis algorithm is used to analyze its periodicity and trend changes, and the flow size in different time periods is predicted. Based on the switch status data, according to the equipment operation frequency and the number of historical state changes, its flow demand is estimated, and network bandwidth resources are allocated based on the prediction results. According to the real-time requirements, the electrical quantity data and key switch status data transmission are allocated to the frequency band or channel with interference less than the set threshold to obtain a hybrid transmission network.

4. The relay protection control method based on multi-source data stream according to claim 1 is characterized in that: The steps for preprocessing electrical quantity data and switch status data are as follows: The weighted moving average filter is used to smooth the electrical quantity data. The specific formula for the processing is: ; in is the filtered value, is the window size of the moving average, Represents the sequence number of discrete time points in a time series; The switch state data is processed by jitter and a time threshold is set. When the time interval between two adjacent state changes is less than the threshold, it is considered to be a false trigger and the original state value is maintained.

5. The relay protection control method based on multi-source data stream according to claim 1 is characterized in that: The steps for building the improved support vector machine model are: Optimize the kernel function parameters and penalty factors of the basic support vector machine. Define the solution space and search space, obtain the particle position vector and particle velocity vector from the solution space, and the initial position and initial velocity of the particle are randomly generated in the search space; Define the particle's own historical optimal position and the group's historical optimal position to update the particle position vector and particle velocity vector; The particle updates its position vector based on the new velocity ; in, Indicates The particle position vector of the iteration; Each particle compares its current fitness value with its historical optimal fitness value. If the current fitness value is better, it updates its historical optimal position. At the same time, it compares the historical optimal fitness values ​​of all particles in the entire particle group to find the historical optimal position of the group. An iteration termination threshold is set, and the iteration is stopped when the iteration termination threshold is reached. The parameter combination corresponding to the historical optimal position of the group, namely the kernel function parameter and the penalty factor, is output as the optimization parameter of the support vector machine model. The support vector machine model is retrained using the optimized parameters to obtain an improved support vector machine model.

6. The relay protection control method based on multi-source data stream according to claim 1 is characterized in that: The updating process of particle velocity vector is: in, is the number of iterations, Indicates that the inertia weight controls the degree to which the particle inherits the current velocity. and is the learning factor, which adjusts the step size of the particle approaching its own historical optimal position and the group's historical optimal position respectively. and are two random numbers.

7. The relay protection control method based on multi-source data stream according to claim 1 is characterized in that: The steps for obtaining the fault behavior mode are: The comprehensive feature vector is extracted and input into the trained improved support vector machine model. The improved support vector machine model determines which fault behavior mode the comprehensive feature vector sample belongs to through the internal optimized kernel function and penalty factor, and outputs the corresponding results, including normal operation, short circuit fault and equipment overload.

8. A relay protection control system based on multi-source data stream, characterized in that: include: Data acquisition module: used to collect electrical quantity data and switch status data; Data transmission processing module: used to pre-process electrical quantity data and switch status data, and build a hybrid transmission network based on the pre-processed electrical quantity data and switch status data; Optimization fusion module: used to extract electrical quantity and switch state data feature vectors from the hybrid transmission network, and fuse the electrical quantity and switch state data feature vectors into a comprehensive feature vector; Pattern acquisition module: used to take the comprehensive feature vector as the input of the improved support vector machine model based on the constructed support vector machine model, and output the fault behavior pattern of the data; Abnormal diagnosis module: used to determine whether the fault behavior mode conforms to the pre-set fault behavior mode and whether the power system has a fault.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the relay protection control method based on multi-source data streams as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of a relay protection control method based on multi-source data streams as described in any one of claims 1 to 7 are implemented.

Images