Relay automatic test system and method based on parallel processing

Abstract

The invention provides an automatic relay testing system and method based on parallel processing, and relates to the technical field of relay testing. The method comprises the following steps: recording a reference signal waveform, time sequence data, a baseline state index and an initial reclosing frequency increment of equipment in a normal operation state; equipment signal difference is monitored and collected in real time in the process of simulating power-off-reclosing, and signal changes are analyzed and sent to the next module; quantifying the system state change to obtain an equipment safety state value; creating a reclosing frequency adaptive algorithm, and judging whether to continue the test and decide the reclosing frequency of the next simulation according to the equipment safety state value and the initial reclosing frequency increment; and according to a test result, determining the maximum tolerable reclosing frequency to obtain a recovery robustness index. The problem that the fixed step length test cannot fit the actual state of the equipment is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of relay testing, and particularly to a relay automatic testing system and method based on parallel processing. Background Art

[0002] Relay protection devices are core components for the safe and stable operation of power systems, and their reliability directly affects the effect of power grid fault isolation and equipment protection. With the expansion of the scale of power systems and the access of new power equipment, relay protection devices face more complex working conditions and higher reliability requirements. After the device experiences a disconnection and reconnection, sensors or instruments may lose their states or be re-initialized, affecting system stability.

[0003] In traditional testing, hand-held testing instruments or single-channel automated systems are used to sequentially perform static and dynamic tests on relays to verify their action characteristics and durability. However, these methods mostly rely on manual operations or simple script driving, and have problems such as low efficiency, susceptibility to human errors, and poor repeatability. In the on-site environment, maintenance personnel often need to connect wires one by one and perform multiple rounds of tests, which is time-consuming and laborious and requires high skills from the staff; test data are mostly scattered and recorded, and subsequent analysis is cumbersome, making it difficult to form comparisons and statistical analyses between batch samples.

[0004] Currently, most relay tests focus on static action characteristics and predefined durability cycles, lacking health perception, and unable to adjust the subsequent test intensity in a timely manner based on the responses of the relay in the previous reclosing operations, which is likely to result in two extremes: one is that the increment is too small, resulting in too many test rounds and long time consumption; the other is that the increment is too large, and once it approaches the critical point, sudden failure may occur, making it impossible to accurately locate the maximum tolerable number of times and increasing the risk of equipment damage. Summary of the Invention

[0005] Object of the Invention: To propose a relay automatic testing system and method based on parallel processing to solve the existing problems mentioned in the above background art.

[0006] To achieve the above object, the present invention provides the following technical solution: A relay automatic testing system based on parallel processing, comprising: An initial state acquisition module, configured to record the baseline signal and the initial reclosing times increment under the normal operating state of the device; A real-time signal monitoring module, configured to monitor and acquire device signals in real time during the simulated power-off - reclosing process, analyze the signal changes in combination with the baseline signal, and send them to the state index extraction module; A state index extraction module, configured to calculate the device safety state value; The autoreclosure times adaptive scheduling module is used to create an autoreclosure times adaptive algorithm. The autoreclosure times adaptive algorithm is used to output the autoreclosure times increment of each relay in the next round according to the initial autoreclosure times increment and the device safety status value, and determine whether to continue the test. If so, it returns to the real-time signal monitoring module; otherwise, it runs the recovery robustness evaluation module. The recovery robustness evaluation module is used to determine the maximum tolerable autoreclosure times according to the real-time device safety status value, and obtain the recovery robustness index.

[0007] A further improvement of the present invention lies in that the initial state acquisition module collects the synchronous trigger time, sampling clock, and waveform signal of each device through a synchronous sampling device, and obtains the baseline characteristics of all data and sends them to the real-time signal monitoring module. Before the start of the test, the initial state acquisition module records and initializes an initial autoreclosure times increment and sends it to the autoreclosure times adaptive scheduling module.

[0008] A further improvement of the present invention lies in that before each simulation event, the real-time signal monitoring module triggers an autoreclosure event through a control terminal, disconnects and then closes the autoreclosure after a set time. After reconnecting the power supply, it continues to record the device output until the signal becomes stable again; records each sampling timestamp, the peak voltage of each phase, and the waveform before and after the power failure; and performs differential analysis on the baseline characteristics corresponding to all data after each simulation and transfers them to the next module.

[0009] A further improvement of the present invention lies in that the autoreclosure times adaptive scheduling module includes a state trend judgment unit and an adaptive increment adjustment unit. After each test ends, the state trend judgment unit obtains the health state drift amount according to the change amount of the real-time device safety status value and combines the latest health state drift amount with the recent M historical health state drift amounts to form a health state drift sequence; counts the number of positive values N+ and the number of negative values N- in the health state drift sequence. If the number of positive values N+ is greater than 70% of all data in the sequence, it is determined that the health state is in a downward trend. If the number of negative values N- is greater than 70% of all data in the sequence, it is determined that the health state is in an upward trend. Otherwise, it is determined that the health state is oscillating.

[0010] A further improvement of the present invention lies in that the adaptive increment adjustment unit is used to run the autoreclosure times adaptive algorithm, and the autoreclosure times adaptive algorithm includes: Define the optimization objective, and define that the particle dimension consists of two components including the current autoreclosure times increment and the steady-state time of this round of testing ; Set the number of particles P. Initially, each particle is uniformly distributed within the domain, and the domain is represented as and , where 、 、 、 respectively represent the minimum reclosing times increment, the maximum reclosing times increment, the minimum test steady-state time, and the maximum test steady-state time set by the system; When adjusting the reclosing times of the device according to the current reclosing times increment and the steady-state time of this round of testing , update the particle velocity and particle position according to the particle update strategy. After evaluating the fitness of each particle according to the objective function, update the individual optimal position and the global optimal position. When the fitness of the individual optimal position of a particle is greater than the fitness of the initial global optimal position, update the individual optimal position of this particle as the new global optimal position, and then output this global optimal position directly as the reclosing times increment of each relay in the next round.

[0011] A further improvement of the present invention lies in that the objective function is expressed as , where represents the negative impact of the health state drift. When the health state is in a downward trend, take the health state drift amount corresponding to the maximum positive value in the health state drift sequence as the negative impact of the health state. When the health state is in an upward trend, the negative impact of the health state drift is taken as 0. When the health state is oscillating, take the average value of the health state drift amount as the negative impact of the health state; represents the test duration, which is expressed as , where represents the average time of one reclosing; represents the maintenance cost, and take , represents the set wear coefficient, 、 and represent the weight coefficients.

[0012] A further improvement of the present invention lies in that the particle update strategy includes: Update the particle velocity and particle position according to the health state drift amount and the health state. Set the initial position of the particle and the velocity , define the individual optimal position , and the global optimal position g is the position corresponding to the minimum value of the objective function among all current ; For each particle i = 1, 2, ..., PIn the k-th iteration, the following rules are executed: extract the health state drift amount at the current moment and the health state drift amount at the previous moment in the health state drift sequence to achieve adaptive inertia weight for updating the inertia weight; for the current particle, first calculate the new trial speed according to the speed of the current particle in the previous round, the historical optimal position and the global current optimal position, in combination with the adaptive inertia weight and the learning factor; after obtaining the new speed, the current position of the particle can be updated.

[0013] A further improvement of the present invention lies in that the adaptive inertia weight specifically includes setting an inertia weight interval , and the adaptive adjustment formula for the inertia weight is: ; where represents the maximum value of the health state drift amount.

[0014] A further improvement of the present invention lies in that after each test, the state index extraction module calculates the device safety state value according to the real-time measured state indexes, specifically including: obtain the synchronization error through the average value of the absolute value of the difference between the synchronization trigger time recorded by the device and its baseline feature; obtain the sampling clock drift through the cumulative offset of the device sampling clock compared with the baseline before and after reclosing; obtain the state retention rate through the proportion of the normal waveform signal observed after reclosing is restored; normalize and sum the synchronization error, the sampling clock drift and the state retention rate with weights to obtain the device safety state value SSI.

[0015] A further improvement of the present invention lies in that during the test, the recovery robustness evaluation module continuously executes the fault-reclosing cycle for each relay sample, and in each round, compares the current device safety state value with the set device safety threshold. When the current device safety state value is greater than the device safety threshold, record the number of reclosing times completed in the previous round as the maximum tolerable reclosing times of the sample , and map to the interval [0, 1] through normalization, denoted as the recovery robustness index.

[0016] On the other hand, the present invention provides a relay automation test method based on parallel processing, including the following steps: S1. Record the baseline signal and the initial reclosing times increment in the normal operation state of the device; S2. During the simulated power-off and reclosing process, monitor and collect the device signals in real time, analyze the signal changes in combination with the baseline signal, and send them to S3; S3. Calculate the device safety state value; S4. Create an adaptive reclosing times algorithm, which is used to output the reclosing times increment of each relay in the next round according to the initial reclosing times increment and the device safety status value, and determine whether to continue the test. If so, return to S2; otherwise, execute S5. S5. Determine the maximum tolerable reclosing times based on the real-time device safety status value to obtain the recovery robustness index.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) By driving the objective function with the health state drift amount and dynamically adjusting the reclosing increment and steady-state waiting time in combination with the particle swarm optimization idea, the present invention can rapidly expand the test step size and accelerate approaching the limit when the health is good; and tighten the step size in time to avoid excessive stress when the health deteriorates. The negative impact assignment for the descending, ascending, and oscillating trends is processed in segments, making the scheduling decision both agile and robust, and balancing the test speed and the protection of the device life.

[0018] (2) In each round of testing, the device safety status value is compared with the threshold in real time. Once the SSI drops below the safety threshold or abnormal opening / closing occurs, the test can be stopped immediately to avoid excessive wear or ablation of the relay contacts; compared with the fixed large-number testing, this system can intelligently terminate in advance, significantly extending the device life and reducing the maintenance cost. Description of the Drawings

[0019] Figure 1 It is a framework diagram of a relay automatic testing system based on parallel processing in the embodiment.

[0020] Figure 2 It is a framework diagram of the reclosing times adaptive scheduling module in the relay automatic testing system.

[0021] Figure 3 It is a flowchart of a relay automatic testing method based on parallel processing in the embodiment. Detailed Embodiment

[0022] In the following description, a large number of specific details are given to provide a more thorough understanding of the present invention. However, it is obvious to those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, some technical features well known to the art are not described to avoid confusion with the present invention.

[0023] Embodiment 1 Figure 1 A framework diagram of a relay automatic testing system based on parallel processing disclosed in this embodiment is shown, including: The initial state acquisition module is used to record the baseline signal and the initial reclosing times increment under the normal operation state of the device; the baseline signal includes the reference voltage and current waveforms and the baseline clock cycle under the normal operation state of the device. The real-time signal monitoring module is used to monitor and collect device signals in real time during the simulated power-off - reclosing process, analyze the signal changes in combination with the baseline signal, and send them to the state index extraction module. The state index extraction module is used to quantify the system state changes and obtain the device safety state value. The reclosing times adaptive scheduling module is used to create a reclosing times adaptive algorithm, and the reclosing times adaptive algorithm is used to output the reclosing times increment of each relay in the next round according to the initial reclosing times increment and the device safety state value, and judge whether to continue the test. If so, it returns to the real-time signal monitoring module. If not, it runs the recovery robustness evaluation module. The recovery robustness evaluation module is used to determine the maximum tolerable reclosing times according to the real-time device safety state value and obtain the recovery robustness index.

[0024] The initial state acquisition module collects the synchronous trigger frames of each device through a synchronous sampling device, and the trigger moment is marked by the transmitter, the data sampling timestamp sequence, the voltage and current waveforms, and the digital quantity status. And obtain the baseline characteristics of all data and send them to the real-time signal monitoring module; before the test starts, the initial state acquisition module records and initializes an initial reclosing times increment , which can be set according to the device rated parameters, the first test results or empirical values, and sent to the reclosing times adaptive scheduling module.

[0025] The real-time signal monitoring module includes triggering a reclosing event through a control terminal before each simulation event, closing the reclosing after a set time after disconnection, and continuing to record the device output until the signal is stable again after reconnecting the power supply. The process is represented as the test steady state time; record each sampling timestamp, the peak voltage of each phase and the waveform before and after the power-off; perform differential analysis on the baseline characteristics corresponding to all data after each simulation and transfer them to the next module.

[0026] The reasons for comparing the number of synchronous trigger frames and the time alignment situation are that if frame loss or offset occurs due to power-off, the trigger moment misalignment can be observed; the reason for comparing the sampling timestamp sequence is that if the clock recalibration error is large or the frequency drifts, the change in the timestamp interval can be seen; the reason for comparing the voltage waveforms is to analyze the overshoot, attenuation or distortion during the voltage recovery process after momentary interruption. The real-time monitoring module can directly transfer the original data to the downstream index module.

[0027] The state index extraction module includes calculating the device safety state value based on the real-time measured state index after each test, specifically including: obtaining the synchronization error by averaging the absolute value of the difference between the synchronization trigger time recorded by the device and its baseline characteristics; for example, if the baseline trigger time sequence is , the real-time triggering sequence of the device is , then the synchronization error , where H represents the number of timestamps; this metric reflects the time alignment accuracy of the device and the reference, and an increase indicates that synchronization is impaired.

[0028] The sampling clock drift is obtained by comparing the cumulative offset of the device sampling clock with the baseline before and after reclosing. For example, if the baseline clock period is Ts, after the first round of testing, the difference between the application intervals of the p-th frame and the q-th frame is the sampling clock drift, which is expressed as This indicator measures the degree of deviation of the device's internal clock or sampling period. Increased drift will accumulate and cause timing misalignment.

[0029] The state retention rate is obtained by measuring the proportion of normal signals observed after the reclosing is restored. For example, the collected analog voltage waveform is first band-pass filtered, and the amplitude of the filtered signal is mapped to the 0-1 interval to facilitate subsequent peak determination. The average amplitude of all peaks is calculated from the baseline phase signal. , set the peak judgment threshold to , Indicates the peak threshold coefficient, which can be 0.7-0.9 in this embodiment; in each sampling window, traverse the normalized signal sequence; if the amplitude of a sampling point is greater than the X points before and after it and exceeds , it is determined to be a valid peak. The selection of X should be greater than or equal to the number of sampling points corresponding to half a cycle to prevent false detection. Peak detection is performed in the baseline window and test window of the same length, and the result is and , directly substitute the peak number obtained above into the formula If R is close to 100%, it means that the loss of signal frames during the simulated power outage-reclosing process is very small and the state is maintained well. If R is significantly lower than 100%, it indicates that there are many frame losses or jitters after reclosing, and the state retention ability is reduced.

[0030] The synchronization error, sampling clock drift and state retention rate are normalized and then weighted summed to obtain the equipment safety status value SSI; the equipment safety status value calculation formula is: ,in, 、 、 Represents weight.

[0031] Example 2 Based on the inventive concept of Embodiment 1, this embodiment proposes a specific implementation manner of the autoreclosure times adaptive scheduling module in a relay automation test system based on parallel processing; Figure 2 The framework diagram of the autoreclosure times adaptive scheduling module in this embodiment is shown, which specifically includes a state trend judgment unit and an adaptive increment adjustment unit; The state trend judgment unit includes, after each test ends, obtaining the health state drift amount according to the change amount of the real-time device safety state value ; subtracting the device safety state value at the previous moment from the device safety state value at the next moment; and combining the latest health state drift amount with the recent M historical health state drift amounts to form a health state drift sequence; counting the number of positive values N+ and the number of negative values N- in the health state drift sequence. If the number of positive values N+ is greater than 70% of all the data in the sequence, it is determined that the health state is in a downward trend. If the number of negative values N- is greater than 70% of all the data in the sequence, it is determined that the health state is in an upward trend. Otherwise, it is determined that the health state is oscillating.

[0032] The adaptive increment adjustment unit is used to run the autoreclosure times adaptive algorithm, including defining the optimization goal as quickly seeking the optimal ΔN increment on the premise of minimizing the test cost, test time, and the risk of relay state deterioration, so that the test can approach the maximum tolerable autoreclosure times within the fewest rounds, and defining that the particle dimension consists of two components including the current autoreclosure times increment and the steady-state time of this round of test which is used to control the time required for the device to cool or the signal to stabilize after each autoreclosure; setting the number of particles P, and initially each particle is uniformly distributed within the domain of definition, and the domain of definition is expressed as and where, , , , respectively represent the minimum autoreclosure times increment, the maximum autoreclosure times increment, the minimum test steady-state time, and the maximum test steady-state time set by the system; when adjusting the autoreclosure times of the device according to the current autoreclosure times increment and the steady-state time of this round of test , update the particle velocity and particle position according to the particle update strategy, update the individual optimal position and the global optimal position after evaluating the fitness of each particle according to the objective function. When the fitness of the individual optimal position of a particle is greater than the fitness of the initial global optimal position, update the individual optimal position of this particle as the new global optimal position, and then output this global optimal position directly as the autoreclosure times increment of each relay in the next round.

[0033] The objective function is expressed as where, Indicates the negative impact of health state drift. When the health state shows a downward trend, the health state drift amount corresponding to the maximum positive value in the health state drift sequence is taken as the negative impact of health state drift. When the health state shows an upward trend, the negative impact of health state drift is set to 0. When the health state is oscillating, the average value of the health state drift amount is taken as the negative impact of health state drift; Is denoted as the test duration, , where Denotes the average time of a single reclosing; Denotes the maintenance cost, taking , Denotes the set wear coefficient. The more times, the higher the wear cost, , and Denote the weight coefficients.

[0034] The beneficial effect of the objective function in this implementation is as follows: Maximum protection for the downward trend: When the health state shows a downward trend, the maximum positive value in the health state drift sequence is directly selected as the "negative impact of health state drift", ensuring that the objective function is highly sensitive to the most severe drift fluctuations. This strategy can quickly capture the most unfavorable deviation of the relay during continuous reclosing, timely amplify the response of the algorithm to unstable states, and thus effectively prevent excessive stress or sudden failure of the relay.

[0035] Safety tolerance for the upward trend: When the health state shows an upward trend, the "negative impact of health state drift" is set to zero, indicating that no negative penalty should be imposed on the reclosing increment at this time. This approach can make full use of the window period with good recovery performance of the relay, allowing the algorithm to moderately increase the test intensity while maintaining safety, accelerating the approach to the maximum tolerable number of times, and avoiding premature convergence due to small drifts, thereby improving the overall test efficiency.

[0036] Smoothing treatment for the oscillating trend: When the health state shows an oscillating trend, the "negative impact of health state drift" is calculated by taking the average value of the health state drift sequence, neither overly amplifying occasional fluctuations nor completely ignoring the drift impact. This balanced strategy can smoothly reflect the short-term jitter of the relay state, avoid frequent adjustment of the step size by the adaptive scheduling due to short-term noise, and achieve more stable step size convergence.

[0037] Improve the robustness and stability of the algorithm. Through the above-mentioned segmented processing, the "negative impact of health state drift" is directly linked to the health trend, enabling the objective function to respond quickly when quickly identifying obvious deterioration and remain stable when the device state is stable or slightly jittery, ensuring that the adaptive scheduling module can exhibit good noise immunity and stability under various test environments and signal quality conditions.

[0038] Optimize the balance between test efficiency and equipment safety. The objective function design fully takes into account both the test speed and the protection of equipment life: when the equipment is in good condition, actively expand the test scope to shorten the test cycle; when the equipment is approaching its limit, quickly tighten the increment to reduce unnecessary reclosing cycles and avoid causing additional wear. The entire process realizes the dual optimization of the efficiency and safety of the "system recovery robustness" evaluation of the relay.

[0039] The particle update strategy includes updating the particle velocity and position according to the health state drift amount and health state, and setting the initial particle position and velocity , define the individual optimal position , the global optimal position g is the position corresponding to the minimum value of the objective function among all current ; for each particle i = 1, 2, ..., P In the k-th iteration, execute the following rules: extract the health state drift amount at the current moment and the health state drift amount at the previous moment in the health state drift sequence to achieve adaptive inertia weight, and use it to update the inertia weight; for the current particle, first calculate the new trial velocity according to the velocity of the current particle in the previous round, the historical optimal position, and the global current optimal position, combined with the adaptive inertia weight and learning factor; after obtaining the new velocity, the current position of the particle can be updated.

[0040] The adaptive inertia weight specifically includes setting the inertia weight interval , then the adaptive adjustment formula of the inertia weight is: ; where represents the maximum value of the health state drift amount.

[0041] This strategy makes the particle have a lower weight and enhanced exploration ability when in high risk ; when the health is stable, increase the weight and the convergence speed is faster.

[0042] The recovery robustness evaluation module includes continuously performing fault-reclosing cycles on each relay sample during the test, and comparing the current equipment safety state value with the set equipment safety threshold in each round. When the current equipment safety state value is greater than the equipment safety threshold, record the cumulative number of reclosing operations completed in the previous round as the maximum tolerable number of reclosing operations for this sample , in order to eliminate the differences in different test conditions or equipment specifications, is mapped to the [0,1] interval through normalization, denoted as the recovery robustness index.

[0043] The set values such as the threshold and weight can be set according to the default settings of the present invention or can be set by those skilled in the art themselves.

[0044] Embodiment 3 Figure 3 The flowchart of an automatic relay testing method based on parallel processing according to the present invention is shown. Based on the same inventive concept as Embodiment 1 and Embodiment 2, the present invention provides an automatic relay testing method based on parallel processing, including the following steps: S1. Record the baseline signal and the initial reclosing times increment under the normal operating state of the device; S2. During the simulated power-off and reclosing process, monitor and collect the device signals in real time, combine with the baseline signal, analyze the signal changes, and send them to S3; S3. Quantify the change of the system state to obtain the device safety state value; S4. Create a reclosing times adaptive algorithm, which is used to output the reclosing times increment of each relay in the next round according to the initial reclosing times increment and the device safety state value, and determine whether to continue the test. If so, return to S2; if not, run S5; S5. Determine the maximum tolerable reclosing times according to the real-time device safety state value to obtain the restoration robustness index.

[0045] Those skilled in the art should understand that the embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of an all-hardware embodiment, an all-software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0046] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram can be realized by computer program instructions, and the combination of the flows and / or blocks in the flowchart and / or block diagram can also be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for realizing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0047] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means realizes the functions in the processFigure 1 one process or multiple processes and / or blocks Figure 1 the functions specified in one block or multiple blocks.

[0048] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0049] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the spirit of the present invention and the scope protected by the claims. All of these are within the protection scope of the present invention.

Claims

1. A relay automatic test system based on parallel processing, characterized in that, Including: An initial state acquisition module, configured to record the baseline signal and the initial reclosing times increment under the normal operation state of the device; A real-time signal monitoring module, configured to monitor and acquire device signals in real time during the simulated power-off - reclosing process, analyze the signal changes in combination with the baseline signal, and send them to the state index extraction module; A state index extraction module, configured to calculate the device safety state value; A reclosing times adaptive scheduling module, configured to create a reclosing times adaptive algorithm, where the reclosing times adaptive algorithm is used to output the reclosing times increment of each relay in the next round according to the initial reclosing times increment and the device safety state value, and determine whether to continue the test. If so, return to the real-time signal monitoring module; otherwise, run the recovery robustness evaluation module; A recovery robustness evaluation module, configured to determine the maximum tolerable reclosing times according to the real-time device safety state value, and obtain the recovery robustness index.

2. The relay automation test system based on parallel processing according to claim 1, wherein, The initial state acquisition module acquires the synchronous trigger time, sampling clock, and waveform signal of each device through a synchronous sampling device, and obtains the baseline characteristics of all data and sends them to the real-time signal monitoring module; Before the test starts, the initial state acquisition module records and initializes an initial reclosing times increment , and sends it to the reclosing times adaptive scheduling module.

3. The relay automation test system based on parallel processing according to claim 1, characterized in that Before each simulation event, the real-time signal monitoring module triggers a reclosing event through a control terminal, disconnects and then closes the reclosing after a set time. After reconnecting the power supply, continue to record the device output until the signal stabilizes again; record each sampling timestamp, peak voltage of each phase, and waveform before and after the power-off; perform differential analysis on the baseline characteristics corresponding to all data after each simulation and transfer them to the next module.

4. The relay automation test system based on parallel processing according to claim 1, wherein, The reclosing times adaptive scheduling module includes a state trend judgment unit and an adaptive increment adjustment unit; After each test, the state trend judgment unit obtains the health state drift amount according to the change amount of the real-time device security state value , and combines the latest health state drift amount with the most recent M historical health state drift amounts to form a health state drift sequence; Count the number of positive values N+ and the number of negative values N- in the healthy state drift sequence. If the number of positive values N+ is greater than 70% of all data in the sequence, then judge that the healthy state is in a downward trend. If the number of negative values N- is greater than 70% of all data in the sequence, then judge that the healthy state is in an upward trend. Otherwise, judge that the healthy state is oscillating.

5. The relay automatic test system based on parallel processing according to claim 4, wherein The adaptive increment adjustment unit is configured to run the reclosing times adaptive algorithm, and the reclosing times adaptive algorithm includes: Define the optimization objective and define that the particle dimension consists of two components including the current reclosing times increment and the steady-state time of this round of tests ; Set the number of particles P. Initially, each particle is uniformly distributed within the domain, which is represented as and , where , , , respectively represent the minimum reclosing times increment, the maximum reclosing times increment, the minimum test steady-state time, and the maximum test steady-state time set by the system; When based on the current reclosing times increment and the steady-state time of this round of testing After adjusting the reclosing times of the device, update the particle velocity and particle position according to the particle update strategy. After evaluating the fitness of each particle according to the objective function, update the individual optimal position and the global optimal position. When the fitness of the individual optimal position of a particle is greater than the fitness of the initial global optimal position, update the individual optimal position of this particle as the new global optimal position, and then output this global optimal position directly as the scheduling parameter of each relay in the next round.

6. The relay automatic test system based on parallel processing according to claim 5, wherein The objective function is expressed as , where represents the negative impact of the health state drift. When the health state is in a downward trend, the health state drift amount corresponding to the maximum positive value in the health state drift sequence is taken as the negative impact of the health state. When the health state is in an upward trend, the negative impact of the health state drift is set to 0. When the health state is oscillating, the average value of the health state drift amount is taken as the negative impact of the health state; represents the test duration, which is expressed as , where represents the average time of a single reclosing; represents the maintenance cost, taking , represents the set wear coefficient, , and represent the weight coefficients.

7. The relay automation test system based on parallel processing according to claim 5, wherein The particle update strategy includes: Update the particle velocity and position according to the health state drift amount and health state, and set the initial particle position and velocity , define the individual optimal position , the global optimal position g is the position corresponding to the minimum value of the objective function among all current ; For each particle i = 1, 2,..., P In the k-th iteration, the following rules are executed: extract the health state drift amount at the current moment and the health state drift amount at the previous moment in the health state drift sequence to achieve adaptive inertia weight for updating the inertia weight; for the current particle, first calculate the new trial speed according to the speed of the current particle in the previous round, the historical optimal position and the global current optimal position, combined with the adaptive inertia weight and the learning factor; after obtaining the new speed, the current position of the particle can be updated.

8. The relay automation test system based on parallel processing according to claim 7, wherein The inertial weight adaptation specifically includes setting an inertial weight interval , and the inertial weight self - adaptation adjustment formula is as follows: ； Among them, represents the maximum value of the health state drift amount.

9. The relay automation test system based on parallel processing according to claim 2, characterized in that After each test, the state index extraction module calculates the device safety state value according to the real-time measured state index, specifically including: Obtain the synchronization error through the average value of the absolute value of the difference between the synchronous trigger time recorded by the device and its baseline characteristics; Obtain the sampling clock drift through the cumulative offset of the device sampling clock compared with the baseline before and after the reclosing; Obtain the state retention rate through the proportion of the normal waveform signal observed after the reclosing recovery; Normalize and sum the synchronization error, sampling clock drift, and state retention rate with weights to obtain the device safety state value SSI; During the test, the recovery robustness evaluation module continuously performs fault-reclosing cycles on each relay sample, and in each round, it compares the current device safety status value with the set device safety threshold. When the current device safety status value is greater than the device safety threshold, it records the number of reclosing operations completed in the previous round as the maximum tolerable reclosing times of the sample. , and is mapped to the interval [0, 1] through normalization, which is denoted as the recovery robustness index.

10. A relay automation test method based on parallel processing, characterized in that, Applied to a relay automation test system based on parallel processing as described in any one of claims 1 to 9, characterized by including the following steps: S1. Record the baseline signal and the initial reclosing times increment under the normal operation state of the device; S2. During the simulation of power outage - reclosing process, monitor and collect device signals in real time, analyze the signal changes in combination with the baseline signal, and send them to S3; S3. Calculate the device safety status value; S4. Create a reclosing times adaptive algorithm, which is used to output the reclosing times increment of each relay in the next round according to the initial reclosing times increment and the device safety status value, and determine whether to continue the test. If so, return to S2; otherwise, run S5; S5. Determine the maximum tolerable reclosing times based on the real - time device safety status value to obtain the recovery robustness index.

Images