Online evaluation system of power grid restoration status based on self-learning mechanism

Abstract

The present invention relates to the field of online evaluation technology, and in particular, to an online evaluation system for power grid recovery status based on a self-learning mechanism. It includes a detection and suppression unit, a governance and recovery unit, an evaluation and judgment unit, a voltage and model unit, and the governance and recovery unit obtains a governance plan from a database to recover the detected waveform problem. The effect judgment module of the present invention is used to receive a command to judge the length of the recovery time in the online evaluation module, and the effect judgment module obtains the number of power outages from the historical data in the detection and suppression unit, and compares the obtained number of power outages with the set standard number of power outages. When the number of power outages after the governance using the governance plan is large within the specified time range, and the length of time for the power grid to be restored is judged to be short, it means that the governance plan is weak. In the face of the situation where the governance plan is not effective, it can promote further optimization and improvement of the governance strategy, thereby improving the robustness and stability of the power grid.

Description

Technical Field

[0001] The invention relates to the technical field of online evaluation, and in particular to an online evaluation system for power grid restoration status based on a self-learning mechanism. Background Art

[0002] Online assessment of grid recovery status based on self-learning mechanism refers to the use of automation technology and machine learning algorithms to conduct real-time assessment and monitoring of grid recovery status. By continuously collecting and analyzing grid operation data, it can identify abnormal conditions or potential problems in the grid and respond in a timely manner to ensure the stable operation and rapid recovery of the grid, improve grid operation efficiency, and reduce power outage time. It is an important part of the development of smart grids.

[0003] With the continuous development of science and technology, when people use a large number of electronic devices, the power grid will accumulate harmonics. In order to avoid the accumulated harmonics causing load distortion in the power grid and causing power outages, the harmonics in the power grid are suppressed. Since the harmonic signals cannot be completely eliminated during the suppression process, when harmonic signals exist, they will interact with the fundamental signal to produce complex voltage and current waveform distortion problems, causing power outages. Therefore, the governance plan is used for recovery. However, during the recovery process, the power grid produces voltage fluctuations, which slows down the recovery speed of the power grid and increases the number of power outages, resulting in a decrease in the stability of the power grid operation state. Therefore, we provide an online evaluation system for the power grid recovery state based on a self-learning mechanism. Summary of the invention

[0004] The purpose of the present invention is to provide an online evaluation system for power grid restoration status based on a self-learning mechanism to solve the problems raised in the above background technology.

[0005] To achieve the above-mentioned object, the present invention provides an online evaluation system for power grid restoration status based on a self-learning mechanism, comprising a detection and suppression unit, a governance and restoration unit, an evaluation and judgment unit, and a voltage and model unit;

[0006] The detection and suppression unit uses a harmonic analyzer to detect the presence of harmonic signals in the operation state of the power grid and judge the critical point of the harmonic signal, and suppresses the harmonics according to the judged critical point of the harmonic signal, and then obtains historical data from the database;

[0007] The governance recovery unit is used to detect the data in the suppression unit and detect the waveform distortion problem, and obtain the governance solution from the database to recover the detected waveform problem;

[0008] The evaluation and judgment unit is used to receive data from the governance and recovery unit and perform online evaluation of the power grid recovery time status, and judge the effectiveness of the governance plan based on the evaluated power grid recovery time status and the number of power outages;

[0009] The voltage and model unit is used to receive the strong or weak effect data of the governance plan judged in the evaluation and judgment unit. The voltage and model unit obtains historical data from the detection and suppression unit. The voltage and model unit obtains governance plan data from the governance recovery unit, and establishes a simulation model based on the historical data and governance plan data.

[0010] As a further improvement of the present technical solution, the detection and suppression unit uses a harmonic analyzer to detect the harmonic signals existing in the operation state of the power grid, and uses the detected harmonic signals and the set harmonic signal critical value for judgment. When it is judged that the detected harmonic signal reaches the set critical value, the harmonic suppressor is triggered, and then the harmonic suppressor is used to generate a corresponding reverse signal according to the detected harmonic signal for suppression.

[0011] As a further improvement of the present technical solution, the control and recovery unit is used to receive a command to trigger the harmonic suppressor in the detection and suppression unit, and detect voltage and current waveform distortion problems in the power grid during power grid operation, and then quantify the number of affected users for the detected waveform distortion problem, match control solutions for similar problems in the database based on the detected waveform distortion problem, and the control solution is a newly generated control solution in the most recent time period, and then use the control solution to restore the waveform distortion problem detected in the power grid, and monitor the recovery status of the power grid in real time.

[0012] As a further improvement of the technical solution, the evaluation and judgment unit includes an online evaluation module, a duration effect module and an effect judgment module;

[0013] The online evaluation module is used to receive the quantified number of affected users, the governance plan and the detected waveform distortion problem in the governance recovery unit. The online evaluation module obtains historical data from the detection suppression unit, and inputs the quantified number of affected users, the governance plan, the detected waveform distortion problem and the historical data into a linear regression model. The linear regression model outputs corresponding model parameters and error terms, and then performs an online evaluation of the power grid recovery time status based on the quantified number of affected users, the governance plan, the detected waveform distortion problem and the historical data, and uses the evaluated power grid recovery time and the set standard time to judge the length of the recovery time;

[0014] The duration effect module is used to receive a command from the online evaluation module to determine the length of the recovery duration. When the recovery duration is determined to be long, it indicates that the effect of the governance scheme is weak, and the command indicating the weak effect of the governance scheme is transmitted to the voltage and model unit.

[0015] As a further improvement of the present technical solution, the effect judgment module is used to receive a command in the online evaluation module to judge the length of recovery time. The effect judgment module obtains the number of power outages from the historical data in the detection and suppression unit, and compares the obtained number of power outages with the set standard number of power outages. When the number of power outages that occur after the control plan is used within the specified time range is large, and the power grid recovery time is judged to be short, it means that the control plan is weak, and the command that the control plan is weak is transmitted to the voltage and model unit.

[0016] As a further improvement of the present technical solution, the effect judgment module is used to receive a command in the online evaluation module to judge the length of recovery time. When within the specified time range, the number of power outages after the control plan is used is small, and the time it takes to restore the power grid is judged to be short, it means that the control plan is more effective, and the evaluated power grid recovery time and number of power outages will be displayed.

[0017] As a further improvement of the technical solution, the voltage and model unit includes a fluctuation judgment module, a model learning module and a re-evaluation module;

[0018] The fluctuation judgment module is used to receive commands of insufficient effectiveness in the duration effect module and the effect judgment module. The fluctuation judgment module obtains the real-time monitored power grid recovery status data from the governance and recovery unit, and at the same time judges the voltage fluctuation by setting the voltage fluctuation threshold and the voltage data in the monitored power grid recovery status data. When the voltage data exceeds the set range, it means that voltage fluctuations have been generated in the recovery process of using the governance plan to govern the waveform distortion problem of the power grid, which slows down the recovery duration of the power grid and triggers a threshold alarm.

[0019] As a further improvement of the present technical solution, the model learning module is used to receive historical data obtained in the detection and suppression unit and the governance plan and monitored power grid recovery status data in the governance and recovery unit, and to establish a simulation model based on the historical data and the governance plan, and then input the governance plan, historical data and power grid recovery status data into the simulation model for self-learning, formulate a new governance plan based on the results of self-learning, and then use the simulation model to evaluate the number of power outages of the new governance plan.

[0020] As a further improvement of the present technical solution, the re-evaluation module is used to receive the voltage fluctuation data in the fluctuation judgment module and the new control plan, simulation model, and the number of power outages for evaluating the new control plan in the model learning module, receive the waveform distortion problems and historical data detected in the control recovery unit and the detection suppression unit, and introduce the voltage fluctuation data and the new control plan into the simulation model to simulate the voltage fluctuation situation during the power grid recovery operation and perform corresponding control and adjustment, and then perform a re-online evaluation of the power grid recovery time status based on the new control plan, the detected waveform distortion problems and the historical data, and use the re-evaluated power grid recovery time and the number of power outages for evaluating the new control plan to judge the strong or weak control effect of the new control plan.

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] 1. In the online evaluation system of the power grid recovery status of the self-learning mechanism, the effect judgment module is used to receive the command of judging the length of recovery time in the online evaluation module. The effect judgment module obtains the number of power outages from the historical data in the detection and suppression unit, and compares the obtained number of power outages with the set standard number of power outages. When the number of power outages after the governance plan is used within the specified time range is large, and the power grid recovery time is judged to be short, it means that the governance plan is weak. In the face of poor results of the governance plan, further optimization and improvement of the governance strategy can be promoted to improve the robustness and stability of the power grid.

[0023] 2. In the online evaluation system of the power grid recovery status of the self-learning mechanism, the re-evaluation module introduces the voltage fluctuation data and the new governance plan into the simulation model, simulates the voltage fluctuation during the power grid recovery operation and performs corresponding control and adjustment, and then re-evaluates the power grid recovery time status online based on the new governance plan, the detected waveform distortion problem and historical data. The re-evaluated power grid recovery time and the number of power outages used to evaluate the new governance plan are used to judge the strong or weak governance effect of the new governance plan. Combining the two indicators of power grid recovery time and power outage number can provide a more comprehensive evaluation, which not only considers the recovery speed of the power grid after a fault, but also includes the effect of the governance plan in reducing power outages. The comprehensive evaluation can fully understand the impact of the new governance plan on the overall operational stability of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is an overall block diagram of the present invention;

[0025] Figure 2 This is a block diagram of the evaluation and judgment unit of the present invention;

[0026] Figure 3 It is a block diagram of the voltage and model unit of the present invention.

[0027] The meaning of each number in the figure is:

[0028] 1. Detection and suppression unit; 2. Governance and recovery unit;

[0029] 3. Evaluation and judgment unit; 31. Online evaluation module; 32. Duration effect module; 33. Effect judgment module;

[0030] 4. Voltage and model unit; 41. Fluctuation judgment module; 42. Model learning module;

[0031] 43. Evaluate the module again. DETAILED DESCRIPTION

[0032] The following will be combined with the accompanying drawings in the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] Example 1

[0034] The present invention provides an online evaluation system for power grid recovery status based on a self-learning mechanism. Figure 1-Figure 3 , including a detection and suppression unit 1, a governance and recovery unit 2, an evaluation and judgment unit 3, and a voltage and model unit 4;

[0035] The detection and suppression unit 1 uses a harmonic analyzer to detect the presence of harmonic signals in the operation state of the power grid and judge the critical point of the harmonic signal, and suppresses the harmonics according to the judged critical point of the harmonic signal, and then obtains historical data from the database; the control and recovery unit 2 is used to detect the data in the suppression unit 1 and detect the waveform distortion problem, and obtains the control plan from the database to restore the detected waveform problem; the evaluation and judgment unit 3 is used to receive the data in the control and recovery unit 2 and perform an online evaluation of the power grid recovery time status, and judge the strength of the effect of the control plan according to the evaluated power grid recovery time status and the number of power outages; the voltage and model unit 4 is used to receive the strong or weak effect data of the control plan judged in the evaluation and judgment unit 3, the voltage and model unit 4 obtains historical data from the detection and suppression unit 1, and the voltage and model unit 4 obtains the control plan data from the control and recovery unit 2, and establishes a simulation model according to the historical data and the control plan data.

[0036] The following is a refinement of the above units, see Figure 1-Figure 3 ;

[0037] The detection and suppression unit 1 uses a harmonic analyzer to detect harmonic signals existing in the operation state of the power grid, and obtains historical data at the same time, and uses the detected harmonic signals and the set harmonic signal critical value for judgment. When it is judged that the detected harmonic signal reaches the set critical value, the harmonic suppressor is triggered, and then the harmonic suppressor is used to generate a corresponding reverse signal according to the detected harmonic signal to suppress it. Harmonic analysis and harmonic suppression technology can be effectively used to maintain the stability and quality of the power grid, ensure that the harmonic level of the power grid is within an acceptable range during operation, and reduce adverse effects on equipment and users;

[0038] Historical data includes the number of power outages (the number of power outages refers to the number of power outages that occur within a specified time after the waveform distortion problem is treated using the same treatment solution), historical power grid operation status data and historical power grid recovery status data, and the total number of power outages that occurred after the waveform distortion problem was treated using the same treatment solution;

[0039] The control and recovery unit 2 is used to receive the command for triggering the harmonic suppressor in the detection and suppression unit 1. When harmonic signals exist, they will interact with the fundamental signal to generate complex voltage and current waveform distortion problems, and detect the voltage and current waveform distortion problems in the power grid during the operation of the power grid, and then quantify the number of users affected by the detected waveform distortion problems, and match the control solutions for similar problems in the database according to the detected waveform distortion problems, and the control solutions are newly generated in the most recent time period. The control solutions are then used to restore the waveform distortion problems detected in the power grid, and the recovery status of the power grid is monitored in real time to obtain real-time monitored power grid recovery status data. The control solution can effectively restore the waveform distortion in the power grid, help reduce voltage fluctuations and frequency distortions in the power grid, and thus improve the stability and reliability of the power grid.

[0040] The evaluation and judgment unit 3 includes an online evaluation module 31, a duration effect module 32 and an effect judgment module 33;

[0041] The online evaluation module 31 is used to receive the quantified number of affected users, the governance plan and the detected waveform distortion problem in the governance recovery unit 2. The online evaluation module 31 obtains historical data from the detection suppression unit 1, and inputs the quantified number of affected users, the governance plan, the detected waveform distortion problem and the historical data into the linear regression model. The linear regression model outputs the corresponding model parameters and error terms, and then performs an online evaluation of the power grid recovery time status based on the quantified number of affected users, the governance plan, the detected waveform distortion problem and the historical data. The recovery time is judged by using the assessed power grid recovery time and the set standard time. The governance plan can effectively restore the waveform distortion in the power grid, help reduce the voltage fluctuation and frequency distortion in the power grid, and thus improve the stability and reliability of the power grid.

[0042] Steps to evaluate the grid restoration time status:

[0043] First, the quantified number of affected users IA, the governance plan RS, the detected waveform distortion problem WDT and the historical data HD are obtained, and then the model parameters CS0, CS1, CS2, CS3, CS4 and the error term corresponding to the quantified number of affected users IA, the governance plan RS, the detected waveform distortion problem WDT and the historical data HD are obtained from the linear regression model. The quantified number of affected users IA, the governance plan RS, the detected waveform distortion problem WDT, the historical data HD and the corresponding model parameters CS0, CS1, CS2, CS3, CS4 and the error term EN are used to perform an online evaluation of the power grid recovery time status, and the evaluated power grid recovery time SH is obtained. The specific algorithm formula is:

[0044] SH=CS0+(CS1×WDT)+(CS2×IA)+(CS3×RS)+(CS4+HD)+EN;

[0045] Among them, CS0 is a constant term. This formula is used to perform online evaluation of the power grid recovery time. Waveform distortion may affect the energy efficiency performance of equipment in the power grid. The adoption of appropriate governance solutions can reduce these effects and improve the efficiency and effectiveness of power transmission.

[0046] The duration effect module 32 is used to receive the command of judging the length of the recovery time in the online evaluation module 31. When the recovery time is judged to be long, it means that the treatment scheme is weak, and the command of insufficient effectiveness is transmitted to the voltage and model unit 4;

[0047] The effect judgment module 33 is used to receive the command for judging the length of the recovery time in the online evaluation module 31. The effect judgment module 33 obtains the number of power outages from the historical data in the detection and suppression unit 1, and compares the obtained number of power outages with the set standard number of power outages. If the number of power outages after the governance scheme is used is large within the specified time range, and the length of the power grid recovery is judged to be short, it means that the governance scheme is weak, and the command with insufficient effectiveness is transmitted to the voltage and model unit 4;

[0048] The effect judgment module 33 is used to receive the command for judging the length of recovery time in the online evaluation module 31. When the number of power outages after the control plan is used within the specified time range is small, and the power grid recovery time is judged to be short, it means that the control plan is effective, and the evaluated power grid recovery time and number of power outages will be displayed.

[0049] The voltage and model unit 4 includes a fluctuation judgment module 41, a model learning module 42 and a re-evaluation module 43;

[0050] The fluctuation judgment module 41 is used to receive commands of insufficient effectiveness in the duration effect module 32 and the effect judgment module 33. The fluctuation judgment module 41 obtains the real-time monitored power grid recovery status data from the governance recovery unit 2, and at the same time judges the voltage fluctuation by setting the voltage fluctuation threshold and the voltage data in the monitored power grid recovery status data. When the voltage data exceeds the set range, it means that voltage fluctuations have been generated in the recovery process of using the governance plan to govern the waveform distortion problem of the power grid, which slows down the recovery duration of the power grid and triggers a threshold alarm.

[0051] The model learning module 42 is used to receive the historical data obtained in the detection and suppression unit 1 and the governance plan and the monitored power grid recovery status data in the governance and recovery unit 2, and establish a simulation model based on the historical data and the governance plan, and then input the governance plan, historical data and the power grid recovery status data into the simulation model for self-learning, formulate a new governance plan based on the results of the self-learning, and then use the simulation model to evaluate the number of power outages of the new governance plan;

[0052] Steps to evaluate the number of power outages of new governance solutions using simulation models:

[0053] First, the total number of outages in the historical data is obtained as sum = (1 + 2 + ... + n), and then the total number of outages is used to evaluate the number of outages of the new governance plan, and the number of outages XT of the new governance plan is obtained. The specific algorithm formula is:

[0054] XT = sum / mab;

[0055] Wherein, mab is a power outage indication function, which is used to indicate whether the nth waveform distortion problem causes a power outage. When the nth waveform distortion problem causes a power outage, mab = 1, otherwise mab = 0;

[0056] The re-evaluation module 43 is used to receive the voltage fluctuation data in the fluctuation judgment module 41 and the new control plan, simulation model, and the number of power outages for evaluating the new control plan in the model learning module 42, receive the waveform distortion problems and historical data detected in the control recovery unit 2 and the detection suppression unit 1, and introduce the voltage fluctuation data and the new control plan into the simulation model to simulate the voltage fluctuation situation during the power grid recovery operation and perform corresponding control and adjustment, and then re-evaluate the power grid recovery time status online based on the new control plan, the detected waveform distortion problems and historical data (the re-evaluation is the same as the above-mentioned operation steps for evaluating the power grid recovery time status), and use the re-evaluated power grid recovery time and the number of power outages for evaluating the new control plan to judge the strong or weak control effect of the new control plan.

[0057] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. The online evaluation system of power grid restoration status based on self-learning mechanism is characterized by: It includes a detection and suppression unit (1), a governance and recovery unit (2), an evaluation and judgment unit (3), and a voltage and model unit (4); The detection and suppression unit (1) uses a harmonic analyzer to detect the presence of harmonic signals in the operation state of the power grid and judge the critical point of the harmonic signal, and suppresses the harmonics according to the judged critical point of the harmonic signal, and then obtains historical data from the database; The control and recovery unit (2) is used to detect the data in the suppression unit (1) and detect the waveform distortion problem, and obtain a control solution from the database to restore the detected waveform problem; The evaluation and judgment unit (3) is used to receive the data in the governance and recovery unit (2) and perform an online evaluation of the power grid recovery time status, and to judge the effectiveness of the governance plan based on the evaluated power grid recovery time status and the number of power outages; The voltage and model unit (4) is used to receive the strong or weak effect data of the governance plan judged in the evaluation and judgment unit (3), the voltage and model unit (4) obtains historical data from the detection and suppression unit (1), and the voltage and model unit (4) obtains governance plan data from the governance recovery unit (2), and establishes a simulation model based on the historical data and the governance plan data.

2. The online evaluation system for power grid recovery status based on the self-learning mechanism according to claim 1 is characterized by: The detection and suppression unit (1) uses a harmonic analyzer to detect harmonic signals existing in the operation state of the power grid, and uses the detected harmonic signals and a set harmonic signal critical value for judgment. When it is judged that the detected harmonic signal reaches the set critical value, the harmonic suppressor is triggered, and then the harmonic suppressor generates a corresponding reverse signal according to the detected harmonic signal to suppress it.

3. The online evaluation system for power grid recovery status based on the self-learning mechanism according to claim 2 is characterized by: The control and recovery unit (2) is used to receive a command from the detection and suppression unit (1) to trigger the harmonic suppressor, and detect voltage and current waveform distortion problems in the power grid during power grid operation, and record the time when the waveform distortion problem occurs, and then quantify the number of users affected by the detected waveform distortion problem, match a control solution for a similar problem in a database based on the detected waveform distortion problem, and the control solution is a control solution newly generated in the most recent time period, and then use the control solution to restore the waveform distortion problem detected in the power grid, and monitor the recovery status of the power grid in real time.

4. The online evaluation system for power grid recovery status based on the self-learning mechanism according to claim 3 is characterized by: The evaluation and judgment unit (3) comprises an online evaluation module (31), a duration effect module (32) and an effect judgment module (33); The online evaluation module (31) is used to receive the quantified number of affected users, the governance plan and the detected waveform distortion problem in the governance recovery unit (2). The online evaluation module (31) obtains historical data from the detection suppression unit (1), and inputs the quantified number of affected users, the governance plan, the detected waveform distortion problem and the historical data into a linear regression model. The linear regression model outputs corresponding model parameters and error terms, and then performs an online evaluation of the power grid recovery time status based on the quantified number of affected users, the governance plan, the detected waveform distortion problem and the historical data, and uses the evaluated power grid recovery time and the set standard time to judge the length of the recovery time. The duration effect module (32) is used to receive a command from the online evaluation module (31) to determine whether the recovery duration is long. When the recovery duration is determined to be long, it indicates that the treatment solution is less effective, and the command with insufficient effectiveness is transmitted to the voltage and model unit (4).

5. The online evaluation system for power grid restoration status based on the self-learning mechanism according to claim 4 is characterized by: The effect judgment module (33) is used to receive a command from the online evaluation module (31) to judge whether the recovery time is short. The effect judgment module (33) obtains the number of power outages from the historical data in the detection and suppression unit (1), and compares the obtained number of power outages with the set standard number of power outages. If the number of power outages after the treatment using the treatment plan is large within the specified time range, and the time for the power grid to be restored is short, it means that the treatment plan is not effective, and the command with insufficient effectiveness is transmitted to the voltage and model unit (4).

6. The online evaluation system for power grid restoration status based on the self-learning mechanism according to claim 4 is characterized by: The effect judgment module (33) is used to receive a command from the online evaluation module (31) to judge the duration of recovery. If the number of power outages after the treatment using the treatment plan is small within a specified time range, and the duration of power grid recovery is judged to be short, it means that the treatment plan is effective, and the evaluated power grid recovery duration and number of power outages are displayed.

7. The online evaluation system for power grid recovery status based on the self-learning mechanism according to claim 5 is characterized by: The voltage and model unit (4) comprises a fluctuation determination module (41), a model learning module (42) and a re-evaluation module (43); The fluctuation judgment module (41) is used to receive commands of insufficient effectiveness from the duration effect module (32) and the effect judgment module (33). The fluctuation judgment module (41) obtains the real-time monitored power grid recovery status data from the governance recovery unit (2), and at the same time judges the voltage fluctuation by setting a voltage fluctuation threshold and the voltage data in the monitored power grid recovery status data. When the voltage data exceeds the set range, it indicates that voltage fluctuations have occurred in the recovery process of using the governance solution to govern the waveform distortion problem of the power grid, which slows down the recovery duration of the power grid and triggers a threshold alarm.

8. The online evaluation system for power grid recovery status based on the self-learning mechanism according to claim 7 is characterized by: The model learning module (42) is used to receive the historical data obtained in the detection and suppression unit (1) and the governance plan and monitored power grid recovery status data in the governance and recovery unit (2), and to establish a simulation model based on the historical data and the governance plan, and then input the governance plan, historical data and power grid recovery status data into the simulation model for self-learning, formulate a new governance plan based on the results of the self-learning, and then use the simulation model to evaluate the number of power outages of the new governance plan.

9. The online evaluation system for power grid recovery status with a self-learning mechanism according to claim 8 is characterized by: The re-evaluation module (43) is used to receive the voltage fluctuation data in the fluctuation judgment module (41) and the new control plan, simulation model, and the number of power outages for evaluating the new control plan in the model learning module (42), receive the waveform distortion problems and historical data detected in the control recovery unit (2) and the detection suppression unit (1), and introduce the voltage fluctuation data and the new control plan into the simulation model to simulate the voltage fluctuation during the power grid recovery operation and perform corresponding control and adjustment, and then re-evaluate the power grid recovery time status online based on the new control plan, the detected waveform distortion problems and historical data, and use the re-evaluated power grid recovery time and the number of power outages for evaluating the new control plan to judge the strong or weak control effect of the new control plan.

Images