Multi-channel information fusion relay protection closed loop automatic test system and method

Abstract

The invention discloses a multi-channel information fusion relay protection closed loop automatic test system and method, and the system comprises a relay protection tester, a merging unit connected with the relay protection tester, and a relay protection device connected with the merging unit. The test control terminal performs signal transmission with the relay protection tester and is used for completing relay protection closed-loop automatic test, the relay protection device is connected and communicated with the test control terminal, an optical fiber communication module is arranged in the relay protection device, and the relay protection device is connected with a test module of the relay protection tester through an optical fiber. A wireless communication module is integrated in the relay protection tester. According to the method, the universality is high, the protection regular checking efficiency is improved, the real-time performance is better, the safety of test data is better, the effectiveness, quickness, sensitivity and selectivity indexes of the relay protection device can be accurately and effectively analyzed and judged, and technical support is provided for relay protection work.

Description

Technical Field

[0001] The invention belongs to the technical field of relay protection testing, and in particular relates to a multi-channel information fusion relay protection closed-loop automatic testing system and method. Background Art

[0002] As the most effective and important technical means to ensure the safety of power equipment and the stability of the power system, relay protection shoulders the important responsibility of protecting the safe operation of power grids and power equipment. With the continuous development of the power grid, the supporting relay protection principles are becoming increasingly complex, resulting in a wide variety of relay protection devices and multiple configurations. Relay protection device testing is an important means to check circuit errors and equipment defects before the relay protection device is put into operation. The test method of relay protection information in the existing technology is to use the relay protection tester and the power automation monitoring system master station to manually set the experimental parameters, manually operate the relay protection tester to apply analog quantities, and then manually check the correctness of the information through the power monitoring platform; it has the disadvantages of numerous projects, complex work, high skill requirements, long working hours, low test efficiency, and large influence of human factors.

[0003] Introducing automated tools to perform automated testing during the testing process is a very practical and efficient method. At present, domestic and foreign relay protection tester manufacturers and relay protection device manufacturers are designing and developing automated testing for relay protection; however, the automatic testing equipment in the existing technology is not a strict automatic testing system, and there are still the following defects and deficiencies:

[0004] First, due to the advancement of computer and communication technology, and its deep integration with power automation technology, the past electromechanical and integrated circuit protection equipment has become digital equipment, and integrated with a variety of protection measurement and control functions, and the way protection equipment transmits information has also changed to digital communication mode; these changes have brought many challenges to the production, testing and operation of protection equipment. First, the protection inspection work involves a wide range of areas, a long construction period, and complicated work content; the current maintenance work is mostly comprehensive maintenance, and secondary maintenance is often carried out during the primary maintenance power outage. When the primary work is in progress, the secondary work often cannot be carried out normally. Second, the execution of secondary safety measures, protection belt switch transmission, set value verification and load testing can only be completed manually; to shorten the protection inspection time, we can only explore how to shorten the inspection time of the protection device itself; currently, the principles and functions of protection devices of various manufacturers are basically the same; therefore, it is necessary to increase investment in science and technology, develop automatic modular testing technology, shorten the inspection time of the protection device itself, and thus improve the efficiency of protection inspection;

[0005] Second, since the test end generally uses a PC, each time a test operation is performed, the PC must be connected to the tester first. The two must be debugged and connected successfully before the test operation can be performed. On the one hand, the risk of problems during the test process is increased, and it is also inconvenient to carry a PC for on-site testing. The test report generated after the test is completed is stored in the PC disk, and the test management center cannot view the test results in real time;

[0006] Third, there is only process-layer data interaction between the tester and the equipment. The tester cannot modify the parameters of the equipment such as fixed values ​​to complete the information of the station control layer. In addition, the executable test items in this test scheme are relatively fixed, mainly testing the protection function logic, completing the local test closed loop through time control or receiving the input node, and the multi-function test can only be simply combined through the functional unit module, and it is impossible to design a more complex test process according to the test requirements;

[0007] Fourth, the test instrument software systems in the prior art basically adopt the mode setting of "protection function test program + test control program". The protection function test program completes the test of the specific protection function, and the test control program calls the protection function test program to edit the test parameters and save the parameter files. According to the test parameters, the specific test program is called to control the test process and form a test report in a custom format; it is also impossible to realize the automatic test of the background, tester and equipment communicating with each other in two directions, that is, it is impossible to realize closed-loop automatic testing; the collection method of the automatic test products of each tester manufacturer is relatively simple, and the judgment logic is simple. The degree of automation is often reduced due to the inability to effectively collect information, and manual intervention is required;

[0008] Fifth, most relay protection tests do not have a state recognition function. When faced with different faults, the same measures are often used, which lack effectiveness and specificity, resulting in unsatisfactory protection effects. The relay protection device cannot be directly, effectively, and quickly protected from electromagnetic interference. Summary of the invention

[0009] The technical problem to be solved by the present invention is to provide a multi-channel information fusion relay protection closed-loop automatic test system for the above-mentioned deficiencies in the prior art, which has strong versatility, improves the efficiency of protection inspection, has better real-time performance, and has better security of test data.

[0010] To solve the above technical problems, the technical solution adopted by the present invention is: a multi-channel information fusion relay protection closed-loop automatic test system, including a relay protection tester, a merging unit connected to the relay protection tester and a relay protection device connected to the merging unit, and a test control terminal that transmits signals with the relay protection tester and is used to complete the relay protection closed-loop automatic test, the relay protection device is connected to and communicates with the test control terminal, an optical fiber communication module is arranged in the relay protection device, and is connected to the test module of the relay protection tester via optical fiber, and a wireless communication module is integrated in the relay protection tester.

[0011] In the above-mentioned multi-channel information fusion relay protection closed-loop automatic test system, the test control terminal is the industrial computer that comes with the relay protection tester, the relay protection tester is connected to the industrial computer through a switch, and the relay protection device is connected to the industrial computer through a network cable.

[0012] In the above-mentioned multi-channel information fusion relay protection closed-loop automatic testing system and method, the test control terminal is a control module integrated inside the relay protection tester. In the relay protection tester, the control module and the test module are connected via a data cable, and the relay protection device is connected to the control module of the relay protection tester via a network cable.

[0013] In the above-mentioned multi-channel information fusion relay protection closed-loop automatic test system, the wireless communication module is a Bluetooth module.

[0014] In the above-mentioned multi-channel information fusion relay protection closed-loop automatic test system, the relay protection tester is integrated with a voltage detection circuit for detecting the voltage of the relay protection device.

[0015] The present invention also discloses a multi-channel information fusion relay protection closed-loop automatic testing method, which can accurately and effectively analyze and judge the effectiveness, speed, sensitivity and selectivity indicators of the relay protection device without manual intervention, and provide technical support for relay protection work; the method comprises the following steps:

[0016] Step 1: Develop a test plan for the relay protection device on the test control terminal, design a test template, including the standard test process for the relay protection device under test, the test method for each test item, and the test result judgment method; set a report template, associate and bind the parameter data and result data in the test template with the report document location, and realize the function of automatically writing the parameter data and result data in the test template into the report template;

[0017] Step 2: Develop a test instrument interface module and an automatic test module on the relay protection tester. The test instrument interface module is compatible with the conventional station relay protection device interface and the digital station relay protection device interface. The automatic test module automatically executes the test items in the test plan according to the test plan, automatically determines whether the test results are qualified, and saves the test results to the set report template;

[0018] Step 3: Perform relay protection closed-loop automation test. The specific process is as follows:

[0019] Step 301: Input a start automatic test command on the mobile terminal and send it wirelessly to the test control terminal; or directly input a start automatic test command on the test control terminal to start the relay protection closed-loop automatic test;

[0020] Step 302: The test control terminal communicates with the relay protection device via MMS, and reads data from the relay protection device or modifies data of the relay protection device. The data read from the relay protection device include set values, pressure plates, measured values, device parameters, device action information, and alarm information. The data modified from the relay protection device include device parameters, set values, and pressure plates.

[0021] Step 303, the test control terminal transmits the protection test function mark and protection test function parameter data to the test module of the relay protection tester, the relay protection tester applies the analog quantity to the merging unit, the merging unit acts on the relay protection device, starts the test, and waits for the return of the test end information. After receiving the test end information, the test result is read and it is determined whether the test result is qualified. When the test result is qualified, the test result data is automatically written into the report template; when the test result is unqualified, it returns to retest.

[0022] In the above method, the relay protection tester is integrated with a voltage detection circuit for detecting the voltage of the relay protection device;

[0023] When designing the test template in step 1, it also includes designing a relay protection device state recognition model based on the Tensorflow neural network. The specific process is as follows:

[0024] Step 101, collecting voltage data of the relay protection device in a normal state, a lightning strike state, a switch operation state, and a short circuit state to form training samples;

[0025] Step 102: Perform Fourier transform on the voltage data to obtain a frequency domain signal, perform feature extraction on the frequency domain signal, and extract the average value, standard deviation, maximum value, minimum value, and root mean square value of the voltage data as a state recognition feature vector;

[0026] Step 103: using the state recognition feature vector as an input item and setting it as the input layer of the Tensorflow neural network; using the state of the relay protection device as an output item and setting it as the output layer of the Tensorflow neural network; connecting two or more hidden layers between the input layer and the output layer; wherein M neurons are set in each hidden layer; setting an activation function after the neuron; adding a dropout layer after the activation function; adopting a fully connected architecture; and establishing a relay protection device state recognition model based on the Tensorflow neural network;

[0027] Step 104: using improved SSA to optimize the number of neurons in each hidden layer of the Tensorflow neural network, setting a loss function, and using training samples to train the relay protection device state recognition model based on the Tensorflow neural network to obtain the relay protection device state recognition model based on the Tensorflow neural network;

[0028] When the relay protection closed-loop automatic test is performed in step 3, after step 303, the following steps are also included:

[0029] Step 304: The relay protection tester detects the voltage of the relay protection device, performs Fourier transform on the voltage data to obtain a frequency domain signal, performs feature extraction on the frequency domain signal, and extracts the mean value, standard deviation, maximum value, minimum value and root mean square value of the voltage data as a state recognition feature vector; the state recognition feature vector is input into a relay protection device state recognition model based on a Tensorflow neural network to obtain the state of the relay protection device.

[0030] In the above method, the specific process of using the improved SSA algorithm to optimize the number of neurons in each hidden layer of the Tensorflow neural network in step 104 is:

[0031] Step 1041, initializing the sparrow population in the improved SSA algorithm;

[0032] Step 1042, calculating the fitness value of the initial population;

[0033] Step 1043: Update the finder position, joiner position and scout position, and use the improved finder position update formula when updating the finder position. Where t is the current iteration number, t+1 is the next iteration number of the current iteration number, j=1, 2, 3, ..., d, d represents the dimension of the variable to be optimized; i=1, 2, 3, ..., n, n is the number of sparrows of the discoverer, is the j-th dimension position of the i-th discoverer sparrow after iteration t, is the j-th dimension position of the i-th discoverer sparrow at iteration t+1, ω is the weight, R is the map factor, s is the alarm value, S is the safety threshold, is the optimal position of the finder in the tth iteration, Q is a random number that follows a normal distribution, and L is a d-dimensional unit vector;

[0034] Step 1044: Obtain the current updated position. When the new position is better than the old position in the previous iteration, perform the update operation. Otherwise, do not perform the update operation, and continue to perform the iteration operation until the iteration termination condition is met, and finally obtain the optimal position from the global.

[0035] Step 1045: The global optimal position obtained by the improved SSA algorithm is used as the number of neurons M in each hidden layer of the Tensorflow neural network.

[0036] In the above method, the activation function is the LeakyReLU activation function.

[0037] In the above method, the loss function is Among them, y Softmax is the multi-classification cross entropy loss function, y Hinge is the hinge loss function.

[0038] Compared with the prior art, the present invention has the following advantages:

[0039] 1. The multi-channel information fusion relay protection closed-loop automatic test system of the present invention realizes mutual two-way communication among the background, relay protection tester and relay protection device, which helps to realize complete automation of testing, evaluation, report generation and storage without human intervention, and helps to standardize the test process. By setting the electrical connection method and the optical fiber connection method, it can realize the recovery combined with electrical port recovery and optical port recovery, and effectively integrate this information to form a comprehensive judgment criterion. Without human intervention, it can accurately and effectively analyze and judge the effectiveness, speed, sensitivity and selectivity indicators of the relay protection device, and provide technical support for relay protection work.

[0040] 2. The multi-channel information fusion relay protection closed-loop automatic test system and method of the present invention can perform closed-loop automatic tests on conventional station relay protection devices, as well as closed-loop automatic tests on digital station relay protection devices. It has strong versatility, shortens the inspection time of the protection device itself, and improves the efficiency of protection inspection. At the same time, it combines the digital and analog signal outputs of the tester itself to comprehensively analyze the test results, thereby improving the reliability of the test.

[0041] 3. The multi-channel information fusion relay protection closed-loop automatic test system of the present invention adopts the industrial computer of the relay protection tester as the test control terminal, or adopts the relay protection tester as the test control terminal, and effectively integrates the test control terminal and the test instrument, so that the whole system is more integrated and efficient; there is no need to use a PC separately, and there is no need to connect and debug the PC with the tester before each test operation, which is more convenient to use. Moreover, through the wireless communication module, wireless transmission, remote storage and viewing of test results can be realized, the real-time performance is better, and the security of test data is also better.

[0042] 4. The present invention designs electrical connections, optical fiber connections, and communication methods to enable multiple data interactions between the process layer and the station control layer between the relay protection tester and the relay protection device. It can also design more complex test processes according to test requirements. It is fully functional and easy to use for secondary development.

[0043] 5. The multi-channel information fusion relay protection closed-loop automated testing method of the present invention adopts a hierarchical structure and modular design concept, effectively isolates the automatic testing part, the report wireless transmission part, and the job form management part, and integrates them according to their functions, making the data flow clearer and the data management more convenient and standardized.

[0044] 6. The multi-channel information fusion relay protection closed-loop automated testing method of the present invention adds the relay protection device status recognition based on the Tensorflow neural network, which helps to adopt targeted measures in the face of different faults, improves the effectiveness and pertinence, can achieve an ideal protection effect, and can directly, effectively and quickly protect the relay protection device from electromagnetic interference, avoiding damage to the equipment.

[0045] 7. The multi-channel information fusion relay protection closed-loop automated testing method of the present invention adopts improved SSA to optimize the number of neurons in each hidden layer of the Tensorflow neural network, and adopts an improved loss function, which helps to improve the state recognition accuracy of the Tensorflow neural network and obtain accurate recognition results.

[0046] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram of a relay protection closed-loop automatic test system with multi-channel information fusion in Example 1 of the present invention;

[0048] Figure 2 It is a schematic diagram of a relay protection closed-loop automatic test system with multi-channel information fusion in Example 2 of the present invention;

[0049] Description of reference numerals:

[0050] 1—Relay protection tester; 2—Merging unit; 3—Relay protection device. DETAILED DESCRIPTION

[0051] Example 1

[0052] like Figure 1 As shown, the multi-channel information fusion relay protection closed-loop automatic test system of this embodiment includes a relay protection tester 1, a merging unit 2 connected to the relay protection tester 1 and a relay protection device 3 connected to the merging unit 2, and a test control terminal that transmits signals with the relay protection tester 1 and is used to complete the relay protection closed-loop automatic test. The relay protection device 3 is connected and communicates with the test control terminal. An optical fiber communication module is arranged in the relay protection device 3, and is connected to the test module of the relay protection tester 1 through an optical fiber to realize data interaction between the test module and the relay protection device 3 under test; a wireless communication module is integrated in the relay protection tester 1 for sending wireless signals to realize wireless data transmission between the mobile terminal and the test control terminal.

[0053] The above hardware structure and connection method of the relay protection closed-loop automatic test system of the present invention realize the mutual two-way communication among the background, the relay protection tester 1, and the relay protection device 3, which is helpful to realize the complete automation of testing, evaluation, report generation and storage without human intervention, and helps to standardize the test process. By setting the electrical connection method and the optical fiber connection method, it is possible to realize the recovery combined with the electrical port recovery and the optical port recovery, and effectively integrate this information to form a comprehensive judgment criterion. Without human intervention, it can accurately and effectively analyze and judge the effectiveness, speed, sensitivity, and selectivity indicators of the relay protection device 3, providing technical support for relay protection work.

[0054] In this embodiment, the test control terminal is the industrial computer 4 provided with the relay protection tester 1. The relay protection tester 1 is connected to the industrial computer 4 via a switch, and the relay protection device 3 is connected to the industrial computer 4 via a network cable to achieve communication with the industrial computer 4.

[0055] This scheme adopts the industrial computer 4 provided with the relay protection tester 1 as the test control terminal. The industrial computer 4 stores a scheme template library. By operating the industrial computer 4, the scheme templates stored in the scheme template library can be edited, and automatic matching schemes can be achieved. The industrial computer 4 sends an automatic test command to the test module of the relay protection tester 1, and the relay protection tester 1 applies an analog quantity to the merging unit 2. The merging unit 2 acts on the relay protection device 3 to automatically complete test-related work.

[0056] In this embodiment, the wireless communication module is a Bluetooth module. The relay protection tester 1 is integrated with a voltage detection circuit for detecting the voltage of the relay protection device 3 .

[0057] Example 2

[0058] like Figure 2 As shown, the multi-channel information fusion relay protection closed-loop automatic test system of this embodiment is different from that of Embodiment 1 in that: the test control terminal is a control module integrated in the relay protection tester 1, in which the control module and the test module are connected via a data line, and the relay protection device 3 is connected to the control module of the relay protection tester 1 via a network cable to achieve communication with the control module of the relay protection tester 1. The rest of the structures are the same as those of Embodiment 1.

[0059] In this solution, the relay protection tester 1 is used as the test control terminal, and the relay protection tester 1 completes the test related work.

[0060] Example 3

[0061] The multi-channel information fusion relay protection closed-loop automatic testing method of this embodiment includes the following steps:

[0062] Step 1: Develop a test plan for the relay protection device 3 on the test control terminal, design a test template, including a standard test process for the relay protection device 3 under test, a test method for each test item, and a test result judgment method; and set a report template, associate and bind the parameter data and result data in the test template with the report document position, and realize the function of automatically writing the parameter data and result data in the test template into the report template;

[0063] Step 2: Develop a test instrument interface module and an automatic test module on the relay protection tester 1. The test instrument interface module is compatible with the interfaces of the conventional station relay protection device 3 and the digital station relay protection device 3. The automatic test module automatically executes the test items in the test plan according to the test plan, automatically determines whether the test results are qualified, and saves the test results to the set report template;

[0064] Step 3: Perform relay protection closed-loop automation test. The specific process is as follows:

[0065] Step 301: Input a start automatic test command on the mobile terminal and send it wirelessly to the test control terminal; or directly input a start automatic test command on the test control terminal to start the relay protection closed-loop automatic test;

[0066] Step 302: the test control terminal communicates with the relay protection device 3 via MMS, and reads data from the relay protection device 3 or modifies the data of the relay protection device 3. The data read from the relay protection device 3 include fixed values, pressure plates, measured values, device parameters, device action information, and alarm information. The data modified from the relay protection device 3 include device parameters, fixed values, and pressure plates.

[0067] Step 303, the test control terminal transmits the protection test function mark and protection test function parameter data to the test module of the relay protection tester 1, the relay protection tester 1 applies the analog quantity to the merging unit 2, the merging unit 2 acts on the relay protection device 3, starts the test, and waits for the return of the test end information. After receiving the test end information, the test result is read and it is determined whether the test result is qualified. When the test result is qualified, the test result data is automatically written into the report template; when the test result is unqualified, it returns to retest.

[0068] During specific implementation, the protection device model can be obtained through multiple channels according to actual conditions. The protection device model can be obtained by parsing SCD and ICD, or by directly connecting the protection device through the station control layer, thus achieving diversification in model generation.

[0069] During the specific implementation, a test record file is also generated to record the test details of each project, including test time, number of tests, test parameter data and result data, etc.

[0070] In this embodiment, the relay protection tester 1 is integrated with a voltage detection circuit for detecting the voltage of the relay protection device 3;

[0071] When designing the test template in step 1, it also includes designing a state recognition model of the relay protection device 3 based on the Tensorflow neural network. The specific process is as follows:

[0072] Step 101, collecting voltage data of the relay protection device 3 in a normal state, a lightning strike state, a switch operation state and a short circuit state to form training samples;

[0073] Step 102: Perform Fourier transform on the voltage data to obtain a frequency domain signal, perform feature extraction on the frequency domain signal, and extract the average value, standard deviation, maximum value, minimum value, and root mean square value of the voltage data as a state recognition feature vector;

[0074] Step 103, taking the state recognition feature vector as an input item, setting it as the input layer of the Tensorflow neural network, taking the state of the relay protection device 3 as an output item, setting it as the output layer of the Tensorflow neural network, connecting two or more hidden layers between the input layer and the output layer, wherein M neurons are set in each hidden layer, an activation function is set after the neuron, and a dropout layer is added after the activation function, and a fully connected architecture is adopted to establish a state recognition model of the relay protection device 3 based on the Tensorflow neural network;

[0075] Step 104: using improved SSA to optimize the number of neurons in each hidden layer of the Tensorflow neural network, setting a loss function, and using training samples to train the relay protection device 3 state recognition model based on the Tensorflow neural network to obtain the relay protection device 3 state recognition model based on the Tensorflow neural network;

[0076] When the relay protection closed-loop automatic test is performed in step 3, after step 303, the following steps are also included:

[0077] Step 304, the relay protection tester 1 detects the voltage of the relay protection device 3, performs Fourier transform on the voltage data to obtain a frequency domain signal, performs feature extraction on the frequency domain signal, and extracts the mean value, standard deviation, maximum value, minimum value and root mean square value of the voltage data as a state recognition feature vector; the state recognition feature vector is input into the state recognition model of the relay protection device 3 based on the Tensorflow neural network to obtain the state of the relay protection device 3.

[0078] In this embodiment, the specific process of using the improved SSA algorithm to optimize the number of neurons in each hidden layer of the Tensorflow neural network in step 104 is:

[0079] Step 1041, initializing the sparrow population in the improved SSA algorithm;

[0080] Step 1042, calculating the fitness value of the initial population;

[0081] Step 1043: Update the finder position, joiner position and scout position, and use the improved finder position update formula when updating the finder position. Where t is the current iteration number, t+1 is the next iteration number of the current iteration number, j=1, 2, 3, ..., d, d represents the dimension of the variable to be optimized; i=1, 2, 3, ..., n, n is the number of sparrows of the discoverer, is the j-th dimension position of the i-th discoverer sparrow after iteration t, is the j-th dimension position of the i-th discoverer sparrow at iteration t+1, ω is the weight, R is the map factor, s is the alarm value, S is the safety threshold, is the optimal position of the finder in the tth iteration, Q is a random number that follows a normal distribution, and L is a d-dimensional unit vector;

[0082] Step 1044: Obtain the current updated position. When the new position is better than the old position in the previous iteration, perform the update operation. Otherwise, do not perform the update operation, and continue to perform the iteration operation until the iteration termination condition is met, and finally obtain the optimal position from the global.

[0083] Step 1045: The global optimal position obtained by the improved SSA algorithm is used as the number of neurons M in each hidden layer of the Tensorflow neural network.

[0084] In the improved SSA algorithm of the present invention, a new finder position update method is proposed, which improves the optimization accuracy and avoids local extreme values. It has the characteristics of good stability, strong global search ability and fast convergence speed. It can optimize the Tensorflow neural network efficiently and accurately, which helps to improve the state recognition accuracy of the Tensorflow neural network, obtain accurate recognition results and avoid misjudgment.

[0085] In this embodiment, the activation function is the LeakyReLU activation function.

[0086] In this embodiment, the loss function is Among them, y Softmax is the multi-classification cross entropy loss function, y Hinge is the hinge loss function. The setting of this loss function helps to improve the training efficiency and the state recognition accuracy of the Tensorflow neural network.

[0087] In order to verify the technical effect of the present invention, experimental verification was carried out. The experimental results show that the present invention can effectively solve the problems of tedious and repetitive testing work and high dependence on laziness of testers in the traditional protection detection mode, avoid missing items due to human factors, reduce the potential safety risks of the device due to incomplete testing, and realize standardized, standardized and efficient closed-loop automatic testing of relay protection.

[0088] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0089] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0090] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0091] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0092] The foregoing description of specific exemplary embodiments of the present invention is for the purpose of illustration and demonstration. These descriptions are not intended to limit the present invention to the precise form disclosed, and it is clear that many changes and variations can be made based on the above teachings. The purpose of selecting and describing the exemplary embodiments is to explain the specific principles of the present invention and its practical application, so that those skilled in the art can realize and utilize various different exemplary embodiments of the present invention and various different selections and changes. The scope of the present invention is intended to be limited by the claims and their equivalents.

Claims

1. A multi-channel information fusion relay protection closed-loop automatic test system, characterized by: The invention comprises a relay protection tester (1), a merging unit (2) connected to the relay protection tester (1), a relay protection device (3) connected to the merging unit (2), and a test control terminal for performing signal transmission with the relay protection tester (1) and for completing relay protection closed-loop automation testing, wherein the relay protection device (3) is connected to and communicates with the test control terminal, an optical fiber communication module is arranged in the relay protection device (3), and is connected to the test module of the relay protection tester (1) via an optical fiber, and a wireless communication module is integrated in the relay protection tester (1).

2. The multi-channel information fusion relay protection closed-loop automatic test system according to claim 1, characterized in that: The test control terminal is an industrial computer (4) provided with the relay protection tester (1); the relay protection tester (1) is connected to the industrial computer (4) via a switch; and the relay protection device (3) is connected to the industrial computer (4) via a network cable.

3. The multi-channel information fusion relay protection closed-loop automatic test system and method according to claim 1, characterized in that: The test control terminal is a control module integrated in the relay protection tester (1). In the relay protection tester (1), the control module and the test module are connected via a data line, and the relay protection device (3) is connected to the control module of the relay protection tester (1) via a network cable.

4. The multi-channel information fusion relay protection closed-loop automatic test system according to claim 1, 2 or 3, characterized in that: The wireless communication module is a Bluetooth module.

5. The multi-channel information fusion relay protection closed-loop automatic test system according to claim 1, 2 or 3, characterized in that: The relay protection tester (1) is integrated with a voltage detection circuit for detecting the voltage of the relay protection device (3).

6. A method for performing multi-channel information fusion relay protection closed-loop automated testing using the relay protection closed-loop automated testing system according to claim 1, 2 or 3, characterized in that: The method comprises the following steps: Step 1: Develop a test plan for the relay protection device (3) on the test control terminal, design a test template, including a standard test process for the relay protection device (3) under test, a test method for each test item, and a test result judgment method; and set a report template, associate and bind parameter data and result data in the test template with the report document location, and realize the function of automatically writing parameter data and result data in the test template into the report template; Step 2: Develop a test instrument interface module and an automatic test module on the relay protection tester (1); the test instrument interface module is compatible with the interface of the conventional station relay protection device (3) and the interface of the digital station relay protection device (3); the automatic test module automatically executes the test items in the test plan according to the test plan, automatically determines whether the test result is qualified, and saves the test result to a set report template; Step 3: Perform relay protection closed-loop automation test. The specific process is as follows: Step 301: Input a start automatic test command on the mobile terminal and send it wirelessly to the test control terminal; or directly input a start automatic test command on the test control terminal to start the relay protection closed-loop automatic test; Step 302: The test control terminal communicates with the relay protection device (3) via MMS, and reads data from the relay protection device (3) or modifies the data of the relay protection device (3). The data read from the relay protection device (3) include set values, pressure plates, measured values, device parameters, device action information, and alarm information. The data modified from the relay protection device (3) include device parameters, set values, and pressure plates. Step 303, the test control terminal transmits the protection test function mark and protection test function parameter data to the test module of the relay protection tester (1), the relay protection tester (1) applies the analog quantity to the merging unit (2), the merging unit (2) acts on the relay protection device (3), starts the test, waits for the return of the test end information, reads the test result after receiving the test end information, and determines whether the test result is qualified. When the test result is qualified, the test result data is automatically written into the report template; when the test result is unqualified, it returns to retest.

7. The method according to claim 6, characterized in that: The relay protection tester (1) is integrated with a voltage detection circuit for detecting the voltage of the relay protection device (3); When designing the test template in step 1, it also includes designing a state recognition model of the relay protection device (3) based on the Tensorflow neural network. The specific process is as follows: Step 101, collecting voltage data of the relay protection device (3) in a normal state, a lightning strike state, a switch operation state and a short circuit state to form training samples; Step 102: Perform Fourier transform on the voltage data to obtain a frequency domain signal, perform feature extraction on the frequency domain signal, and extract the average value, standard deviation, maximum value, minimum value, and root mean square value of the voltage data as a state recognition feature vector; Step 103, taking the state recognition feature vector as an input item and setting it as the input layer of the Tensorflow neural network, taking the state of the relay protection device (3) as an output item and setting it as the output layer of the Tensorflow neural network, connecting two or more hidden layers between the input layer and the output layer, wherein M neurons are set in each hidden layer, an activation function is set after the neuron, and a dropout layer is added after the activation function, and a fully connected architecture is adopted to establish a state recognition model of the relay protection device (3) based on the Tensorflow neural network; Step 104, using improved SSA to optimize the number of neurons in each hidden layer of the Tensorflow neural network, and setting a loss function, using training samples to train the relay protection device (3) state recognition model based on the Tensorflow neural network, to obtain the relay protection device (3) state recognition model based on the Tensorflow neural network; When the relay protection closed-loop automatic test is performed in step 3, after step 303, the following steps are also included: Step 304: the relay protection tester (1) detects the voltage of the relay protection device (3), performs Fourier transform on the voltage data to obtain a frequency domain signal, performs feature extraction on the frequency domain signal, and extracts the mean value, standard deviation, maximum value, minimum value and root mean square value of the voltage data as a state recognition feature vector; The state recognition feature vector is input into a state recognition model of a relay protection device (3) based on a Tensorflow neural network to obtain the state of the relay protection device (3).

8. The method according to claim 7, characterized in that: The specific process of using the improved SSA algorithm to optimize the number of neurons in each hidden layer of the Tensorflow neural network in step 104 is: Step 1041, initializing the sparrow population in the improved SSA algorithm; Step 1042, calculating the fitness value of the initial population; Step 1043: Update the finder position, joiner position and scout position, and use the improved finder position update formula when updating the finder position. Where t is the current iteration number, t+1 is the next iteration number of the current iteration number, j=1, 2, 3, ..., d, d represents the dimension of the variable to be optimized; i=1, 2, 3, ..., n, n is the number of sparrows of the discoverer, is the j-th dimension position of the i-th discoverer sparrow after iteration t, is the j-th dimension position of the i-th discoverer sparrow at iteration t+1, ω is the weight, R is the map factor, s is the alarm value, S is the safety threshold, is the optimal position of the finder in the tth iteration, Q is a random number that follows a normal distribution, and L is a d-dimensional unit vector; Step 1044, obtaining the current updated position, and when the new position is better than the old position of the previous iteration, performing the update operation, otherwise not performing the update operation, and continuing the iteration operation until the iteration termination condition is met, and finally obtaining the optimal position from the global perspective; Step 1045: The global optimal position obtained by the improved SSA algorithm is used as the number of neurons M in each hidden layer of the Tensorflow neural network.

9. The method according to claim 7, characterized in that: The activation function is the LeakyReLU activation function.

10. The method according to claim 7, characterized in that: The loss function is Among them, y Softmax is the multi-classification cross entropy loss function, y Hinge is the hinge loss function.

Images