Power terminal overload decomposition system based on self-repair mechanism

Abstract

The present invention discloses an overload decomposition system for power terminals based on a self-repairing mechanism, which relates to the field of power management technology, including: performing load monitoring, detecting and obtaining an overload parameter sequence when an overload occurs; performing load scale analysis and load scale volatility analysis to obtain load scale information and load volatility information; performing damage probability analysis on overload components in the power terminal, obtaining basic component damage probability and damage probability volatility information, and performing correction to obtain component damage probability; establishing a load reduction parameter space when it is greater than a damage probability threshold, adjusting the load reduction parameter space, and performing load reduction optimization to obtain optimal load reduction parameters, and performing load decomposition on the power terminal. The present invention solves the technical problem that the prior art cannot perform intelligent adjustment on overload of power terminals, realizes intelligent adjustment of power terminals under overload conditions, and achieves the technical effect of reducing component damage risk and improving system stability.

Description

Technical Field

[0001] The present invention relates to the technical field of power management, and in particular to a power terminal overload decomposition system based on a self-repairing mechanism. Background Art

[0002] With the rapid growth of electricity demand and power load, power terminal equipment is increasingly used in power grids. However, as the scale of power systems continues to expand and the complexity of load fluctuations increases, the operation and management of power terminals under high load and overload conditions are facing increasing challenges. Traditional overload protection methods usually prevent equipment damage by simply cutting off power or limiting current, but such methods often lack intelligence and are difficult to meet the needs of modern power systems for efficient, stable and continuous operation of equipment. Summary of the invention

[0003] The present application provides a power terminal overload decomposition system based on a self-repair mechanism, which is used to solve the technical problem that the prior art cannot perform intelligent adjustment for power terminal overload.

[0004] In view of the above problems, the present application provides a power terminal overload decomposition system based on a self-repair mechanism.

[0005] The present application provides a power terminal overload decomposition system based on a self-repair mechanism, the system comprising:

[0006] A load monitoring module, wherein the load monitoring module performs load monitoring in the power terminal, and when overload occurs, continuously detects to obtain an overload parameter sequence; a load analysis module, wherein the load analysis module performs load scale analysis and load scale volatility analysis according to the overload parameter sequence, and obtains load scale information and load volatility information; a damage probability analysis module, wherein the damage probability analysis module performs damage probability analysis of overload components in the power terminal according to the overload parameter sequence, obtains basic component damage probability and damage probability volatility information, and corrects the basic component damage probability to obtain component damage probability; a load decomposition module, wherein when the component damage probability is greater than a damage probability threshold, the load decomposition module establishes a load reduction parameter space according to the load scale information, adjusts the load reduction parameter space according to the load volatility information, performs load reduction optimization, obtains optimal load reduction parameters, and performs load decomposition on the power terminal.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The present application performs load monitoring in the power terminal, and when overload occurs, continuously detects to obtain an overload parameter sequence; based on the overload parameter sequence, performs load scale analysis and load scale volatility analysis to obtain load scale information and load volatility information; based on the overload parameter sequence, performs overload component damage probability analysis in the power terminal to obtain basic component damage probability and damage probability volatility information, and corrects the basic component damage probability to obtain component damage probability; when the component damage probability is greater than the damage probability threshold, a load reduction parameter space is established based on the load scale information, the load reduction parameter space is adjusted based on the load volatility information, and load reduction optimization is performed to obtain the optimal load reduction parameter, and load decomposition is performed on the power terminal. The present invention solves the technical problem that the prior art cannot perform intelligent adjustment for overload of the power terminal, obtains load and damage information through load monitoring and damage probability analysis, constructs and optimizes the load reduction parameter space when overloaded, finds the optimal load adjustment parameter, performs load decomposition, and realizes intelligent adjustment of the power terminal under overload state, so as to reduce the risk of component damage and improve the technical effect of system stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0010] Figure 1 A schematic diagram of the structure of an overload decomposition system for power terminals based on a self-repairing mechanism provided in an embodiment of the present application.

[0011] Figure 2 A schematic flow chart of a method for decomposing overload of a power terminal based on a self-repairing mechanism provided in an embodiment of the present application.

[0012] Explanation of the reference numerals: load monitoring module 11 , load analysis module 12 , damage probability analysis module 13 , load decomposition module 14 . DETAILED DESCRIPTION

[0013] The present application provides an overload decomposition system for power terminals based on a self-repairing mechanism, aiming to solve the technical problem that the prior art cannot perform intelligent adjustment for overload of power terminals. Through load monitoring and damage probability analysis, load and damage information is obtained. When overloaded, the load reduction parameter space is constructed and optimized, the optimal load adjustment parameters are found, and the load is decomposed to achieve intelligent adjustment of the power terminal under overload conditions, thereby achieving the technical effect of reducing the risk of component damage and improving system stability.

[0014] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0015] It should be noted that any variations of the terms "include" and "have" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0016] like Figure 1 As shown, the embodiment of the present application provides a power terminal overload decomposition system based on a self-repair mechanism, which is used to perform the following Figure 2 The power terminal overload decomposition method based on the self-repair mechanism shown in the figure, the system includes:

[0017] The load monitoring module 11 is inside the power terminal and performs load monitoring. When an overload occurs, the load monitoring module 11 continuously detects and obtains an overload parameter sequence.

[0018] In the embodiment of the present application, the load monitoring module is responsible for real-time monitoring of the load status of the power terminal, and when an overload occurs, continuous detection is performed to obtain an overload parameter sequence. The load monitoring module samples load parameters such as current, voltage, and power in real time. When it detects that the load exceeds a preset load threshold, that is, the load level is higher than the upper limit of the safe operation of the power terminal, it determines that it has entered an overload state, and then starts the overload detection process. In this state, the load monitoring module records the load parameters at the current moment, and while the overload state continues, it regularly collects load data to generate an overload parameter sequence containing multiple time points.

[0019] Furthermore, in the system provided by the embodiment of the application, the load monitoring module 11 is also used for:

[0020] In the power terminal, load monitoring is performed to determine whether the load is greater than the load threshold; if so, overload parameter recording is performed, and the overload parameters within the latest N moments are used as the overload parameter sequence; if not, load monitoring is continued, and N is an integer greater than 5.

[0021] In the embodiment of the present application, the load monitoring module monitors the current operating status of the power terminal by collecting key load parameters such as current, voltage and power in real time. The load parameters at each moment are compared with the preset load threshold. The load threshold is a critical value set by technical experts based on the design and safety requirements of the power terminal. When the actual load exceeds this value, the overload processing process will be triggered.

[0022] During the monitoring process, when the load monitoring module determines that the current load parameter exceeds the load threshold, it enters the overload state and starts overload parameter recording. In this state, the load monitoring module will continue to record the load data of the power terminal from that moment on, and collect the load parameters of the latest N moments into an overload parameter sequence. This overload parameter sequence records the load changes of the power terminal under the overload state, reflecting the pressure on the power terminal during the overload period. Here N is an integer greater than 5, which is used to ensure that the number of collected data samples is sufficient.

[0023] On the other hand, if the load monitoring module determines that the load parameter does not exceed the load threshold, it will not enter the overload state, but continue to perform regular load monitoring.

[0024] The load analysis module 12 performs load scale analysis and load scale volatility analysis according to the overload parameter sequence to obtain load scale information and load volatility information.

[0025] In the embodiment of the present application, the load analysis module is responsible for processing the overload parameter sequence to obtain load scale information and load volatility information. Specifically, the load scale information is first obtained by calculating the average value of the overload parameter sequence. Next, the difference between the load parameter and the average value at each time point is analyzed to calculate the fluctuation amplitude relative to the average load at each moment. Then, these fluctuation amplitudes are statistically analyzed to obtain their mean, i.e., the load volatility information.

[0026] Furthermore, in the system provided by the embodiment of the application, the load analysis module 12 is also used for:

[0027] According to the overload parameter sequence, an average overload parameter is calculated to be used as load scale information; according to the overload parameter sequence and the load scale information, load volatility information is calculated to be obtained.

[0028] In an embodiment of the present application, the load analysis module first extracts current, voltage and power data at multiple time points from the overload parameter sequence, and averages these data to obtain an average overload parameter, and uses the calculated average overload parameter as load scale information.

[0029] Next, based on the load scale information obtained, the load volatility is further analyzed. Specifically, each load parameter in the overload parameter sequence is compared with the average overload parameter, that is, the load scale information, and the difference between the load parameter and the average value at each moment is calculated. These differences are then ratioed with the load scale information to obtain the load fluctuation amplitude at each moment, that is, the degree of fluctuation of the load at a certain moment relative to the overall average load. Finally, the load analysis module performs a statistical analysis on all calculated load fluctuation amplitudes, calculates the mean of these fluctuation amplitudes, and generates load volatility information.

[0030] Furthermore, in the system provided by the embodiment of the application, the load analysis module 12 is also used for:

[0031] A first overload parameter is randomly selected from the overload parameter sequence, the difference between the first overload parameter and the load scale information is calculated, and the ratio of the difference to the load scale information is calculated as the first load fluctuation amplitude; the load fluctuation amplitudes of several overload parameters are continuously randomly selected and calculated, the average is calculated, and the load fluctuation information is obtained.

[0032] In an embodiment of the present application, the load analysis module first randomly selects a first overload parameter corresponding to a time point from the overload parameter sequence. After selecting the first overload parameter, the load analysis module compares the parameter with the load scale information calculated previously, and calculates the difference between the two. The above difference is then ratio-calculated with the load scale information, i.e., the average overload parameter, to obtain the first load fluctuation amplitude. This ratio indicates the intensity of load fluctuation at that moment, revealing the relative volatility of the selected overload parameter in the context of the overall load. The larger the load fluctuation amplitude, the more drastic the load change at that moment, which may cause greater pressure on the power terminal.

[0033] In order to obtain more comprehensive load fluctuation information, the load analysis module will not only calculate the fluctuation amplitude of one overload parameter, but will continue to randomly select multiple overload parameters and repeat the above difference calculation and ratio calculation process to obtain multiple load fluctuation amplitudes. After obtaining multiple load fluctuation amplitudes, these fluctuation amplitudes are averaged to obtain load fluctuation information.

[0034] The damage probability analysis module 13 performs damage probability analysis on overload components in the power terminal according to the overload parameter sequence, obtains basic component damage probability and damage probability volatility information, and corrects the basic component damage probability to obtain component damage probability.

[0035] In the embodiment of the present application, the damage probability analysis module first uses the pre-trained model to analyze the overload parameter sequence to predict the basic component damage probability of each component, that is, the average damage risk of the component during the overload period. Then, by randomly selecting a number of damage probability values ​​and comparing them with the basic values, the fluctuation range of the damage risk at different time points, that is, the damage probability volatility information, is obtained. Finally, the basic component damage probability is corrected based on the damage probability volatility information to obtain a more accurate component damage probability.

[0036] Furthermore, in the system provided in the embodiment of the application, the damage probability analysis module 13 is also used for:

[0037] Pre-training an overload component damage predictor of the power terminal; using the overload component damage predictor to predict component damage probabilities for multiple overload parameters in the overload parameter sequence to obtain multiple component damage probabilities; calculating the mean of the multiple component damage probabilities to obtain a basic component damage probability; randomly selecting a number of component damage probabilities from the multiple component damage probabilities, respectively calculating the damage probability fluctuation amplitudes with the basic component damage probability, and calculating the mean to obtain damage probability volatility information; using 1 and the sum of the damage probability volatility information, multiplying it by the basic component damage probability for correction calculation to obtain the component damage probability.

[0038] In an embodiment of the present application, the overload component damage predictor is first pre-trained by using supervised learning, and the training data includes operation records of the same type of power terminals under different overload conditions. The overload component damage predictor of the power terminal is obtained through training.

[0039] After the model training is completed, the overload parameter sequence is input into the pre-trained predictor. The overload parameter sequence contains current, voltage and power data at multiple time points. The predictor uses this data to calculate the component damage probability at each time point through the model to obtain multiple component damage probabilities. After obtaining multiple component damage probabilities, the arithmetic mean method is used to calculate these damage probabilities to obtain the basic component damage probability.

[0040] In order to evaluate the stability of damage risk, several probability values ​​are randomly selected from the component damage probability at multiple time points and compared with the basic component damage probability. Through differential analysis, the damage probability fluctuation amplitude between each selected probability and the basic value is calculated, that is, the absolute value of the difference between the randomly selected component damage probability and the basic component damage probability is divided by the basic component damage probability to obtain the damage probability fluctuation amplitude. After that, the average of multiple fluctuation amplitudes is calculated to obtain the damage probability volatility information, that is, the average value of the fluctuation amplitude after multiple random selections.

[0041] After obtaining the damage probability fluctuation information, the basic component damage probability is corrected by adding 1 to the damage probability fluctuation information and then multiplying it by the basic component damage probability to obtain the component damage probability.

[0042] Furthermore, in the system provided in the embodiment of the application, the damage probability analysis module 13 is also used for:

[0043] According to the overload operation record data of the same model of power terminals, a sample overload parameter set is collected, and the proportion of power terminal component damage under different sample overload parameter operations is collected as a sample component damage probability set; the sample overload parameter set and the sample component damage probability set are used as supervised training data and verification data to train the overload component damage predictor until convergence.

[0044] In the embodiment of the present application, firstly, overload operation record data are extracted from the historical data of multiple power terminals of the same model, and these data cover the operation of the power terminals under different overload conditions. Specifically, the data include parameters such as current, voltage, and power under overload conditions, and these data are generally summarized into a sample overload parameter set.

[0045] While collecting overload parameters, the actual damage of power terminal components under each overload condition is counted. Specifically, the number of times the component is damaged under a certain overload condition is calculated, and the total number of times the equipment is operated under the same overload condition is counted. The damage ratio is obtained by dividing the number of damage events by the corresponding total number of samples. The damage ratio reflects the probability of component damage under a specific overload condition. Through the above process, a set of sample component damage probabilities is obtained.

[0046] Next, the sample overload parameter set is paired with the sample component damage probability set to form the training data set required for supervised learning. In this data set, the overload parameters are used as input features and the component damage probability is used as the output label to guide the training of the predictor. The training data set is input into the overload component damage predictor for training. During the training process, the predictor uses a multi-layer neural network model to gradually adjust the weights and bias parameters in the model so that the predicted damage probability is as close to the actual damage probability as possible. After each prediction, the error between the predicted value and the true value is calculated, and the error size is measured using loss functions such as mean square error. Through the back propagation algorithm, the model updates its internal parameters, gradually reduces the loss function value, and improves the accuracy of the prediction. In order to prevent the model from overfitting on the training data, the training data set is divided into a training set and a validation set in a ratio of 7 to 3. The training set is used to optimize the model parameters, while the validation set is used to test the performance of the model on unseen data. After multiple rounds of training and validation, the predictor gradually optimizes its prediction performance. When the change of the model's loss function on the validation data set tends to be stable and the prediction error reaches the expected target, the model is considered to have converged.

[0047] Through the above process, a model is finally obtained which can predict the probability of component damage according to the overload parameter input, namely, the overload component damage predictor.

[0048] The load decomposition module 14 establishes a load reduction parameter space according to the load scale information when the component damage probability is greater than the damage probability threshold, adjusts the load reduction parameter space according to the load volatility information, performs load reduction optimization, obtains optimal load reduction parameters, and performs load decomposition on the power terminal.

[0049] In an embodiment of the present application, the component damage probability is compared with a preset damage probability threshold. If the component damage probability is greater than the damage probability threshold, the load adjustment process is started. Specifically, first, a preliminary load reduction parameter space is established based on the load scale information, that is, the average load level during the overload period, and the adjustment range is determined. Then, the parameter space is dynamically adjusted in combination with the load volatility information to adapt it to the load changes. Then, an optimization algorithm, such as particle swarm optimization or genetic algorithm, is used to find the optimal load reduction parameters in the parameter space that can most effectively reduce the risk of component damage. Finally, the parameters are applied to load adjust the power terminal to ensure that the equipment returns to a more stable operating state under overload conditions.

[0050] Furthermore, in the system provided by the embodiment of the application, the load decomposition module 14 is also used for:

[0051] Determine whether the component damage probability is greater than the damage probability threshold, if not, do not perform load decomposition; if yes, use the load scale information as the maximum endpoint value of the load reduction parameter space to construct the load reduction parameter space; use 1 and the sum of the load volatility information, multiply by the maximum endpoint value of the load reduction parameter space, adjust to obtain the adjusted load reduction parameter space; in the adjusted load reduction parameter space, randomly generate a first load reduction parameter, and perform component damage probability analysis after load reduction to obtain the first component damage probability, and calculate to obtain the first decomposition fitness; continue to optimize the load reduction parameters in the load reduction parameter space until convergence, output the load reduction parameters with the largest decomposition fitness, obtain the optimal load reduction parameters, and perform load decomposition of load reduction on the power terminal.

[0052] In the embodiment of the present application, the component damage probability is first compared with the preset damage probability threshold. If the damage probability is lower than the threshold, it means that the damage risk of the component is within an acceptable range and the load decomposition process is not started. If the damage probability is higher than the threshold, it means that the overload state may cause serious damage to the component, and load decomposition is required.

[0053] After determining that load decomposition is required, the load scale information is used to construct the load reduction parameter space. Specifically, the load scale information is used as the maximum endpoint value of the parameter space to establish a preliminary adjustment range to ensure that during the load adjustment process, the load will not drop too low and affect the normal operation of the system. For example, if the power in the current overload state is 1000 watts and the load scale information is 900 watts, the maximum endpoint value of the load reduction parameter space is set to 900 watts. The load reduction parameter space is then adjusted using the load volatility information. The result of 1 plus the load volatility information is taken and multiplied by the preliminary maximum endpoint value of the load reduction parameter space to expand the range of the parameter space. Through the above process, the adjusted load reduction parameter space is obtained.

[0054] Then, in the adjusted load reduction parameter space, a first load reduction parameter is randomly generated using the Monte Carlo simulation method. For example, the randomly generated first load reduction parameter is to reduce the power by 50 watts as a preliminary load adjustment plan. After the first load reduction parameter is generated, the load adjustment of the power terminal is simulated based on the component damage prediction model, and the first component damage probability under the parameter is calculated. Then, the first decomposition fitness is calculated by the fitness function.

[0055] After the first load reduction parameters are generated and their fitness is evaluated, the load reduction parameters are further optimized in the adjusted load reduction parameter space through an optimization algorithm, such as particle swarm optimization. Based on the performance of the initial parameters, the optimization algorithm performs multiple iterations and searches in the parameter space to generate new load reduction parameters and gradually improve them to find parameter values ​​with higher decomposition fitness. This process will continue until the parameters that can maximize the reduction of component damage probability are found, and the parameter changes tend to be stable, that is, the convergence state is reached.

[0056] When the optimization process converges, the load reduction parameter with the highest decomposition fitness is selected as the final optimal load reduction parameter. For example, the optimal load reduction parameter finally determined is to reduce the power by 80 watts. Finally, the optimal parameter is applied to the power terminal, and the actual load adjustment operation is performed to decompose the load of the power terminal for load reduction, gradually reducing the original 1000 watt overload state to 920 watts, ensuring that the system returns to a safer and more stable operating state.

[0057] Furthermore, in the system provided by the embodiment of the application, the load decomposition module 14 is also used for:

[0058] The difference between the load scale information and the first load reduction parameter is calculated to obtain the first load parameter after the load reduction, and the difference is input into the overload component damage predictor to obtain the first component damage probability through analysis and prediction; according to the first component damage probability and the first load reduction parameter, the first decomposition fitness is calculated and obtained as follows:

[0059] ;

[0060] Among them, PF is the decomposition fitness, and is the weight, K is the component damage probability, is the load reduction parameter, and PA is the load parameter after load reduction.

[0061] In the embodiment of the present application, the difference between the load scale information and the first load reduction parameter is first calculated to obtain the first load parameter after the load reduction. Then, the first load parameter after the load reduction is input into the overload component damage predictor for analysis, and the first component damage probability under the new load condition is obtained through the prediction model.

[0062] Then, according to the calculated first component damage probability and the first load reduction parameter, the first decomposition fitness is calculated by the fitness function, and the fitness function is: ; Among them, PF is the decomposition fitness, and is the weight, which is pre-set by technical experts, K is the component damage probability, is the load reduction parameter, that is, the load difference before and after adjustment, indicating the amplitude of load adjustment; PA is the load parameter after load reduction, indicating the new load level after load adjustment.

[0063] Through the above process, the first decomposition fitness is obtained. Through fitness evaluation, the effect of the load adjustment scheme on component protection is accurately judged, providing a basis for further optimizing the load adjustment parameters, while avoiding excessive load reduction leading to a decrease in the operating performance of the power terminal.

[0064] In the embodiments of the present application, in summary, the embodiments of the present application have at least the following technical effects:

[0065] The present application performs load monitoring in the power terminal, and when overload occurs, continuously detects to obtain an overload parameter sequence; based on the overload parameter sequence, performs load scale analysis and load scale volatility analysis to obtain load scale information and load volatility information; based on the overload parameter sequence, performs overload component damage probability analysis in the power terminal to obtain basic component damage probability and damage probability volatility information, and corrects the basic component damage probability to obtain component damage probability; when the component damage probability is greater than the damage probability threshold, a load reduction parameter space is established based on the load scale information, the load reduction parameter space is adjusted based on the load volatility information, and load reduction optimization is performed to obtain the optimal load reduction parameter, and load decomposition is performed on the power terminal. The present invention solves the technical problem that the prior art cannot perform intelligent adjustment for overload of the power terminal, obtains load and damage information through load monitoring and damage probability analysis, constructs and optimizes the load reduction parameter space when overloaded, finds the optimal load adjustment parameter, performs load decomposition, and realizes intelligent adjustment of the power terminal under overload state, so as to reduce the risk of component damage and improve the technical effect of system stability.

[0066] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0067] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0068] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. The power terminal overload decomposition system based on the self-repair mechanism is characterized by: The system comprises: A load monitoring module, which performs load monitoring in the power terminal and continuously detects and obtains an overload parameter sequence when an overload occurs; A load analysis module, wherein the load analysis module performs load scale analysis and load scale volatility analysis according to the overload parameter sequence to obtain load scale information and load volatility information; A damage probability analysis module, wherein the damage probability analysis module performs damage probability analysis on overload components in the power terminal according to the overload parameter sequence, obtains basic component damage probability and damage probability volatility information, and corrects the basic component damage probability to obtain component damage probability; A load decomposition module, when the component damage probability is greater than a damage probability threshold, uses the load scale information as the maximum endpoint value of the load reduction parameter space to establish a load reduction parameter space, multiplies the sum of 1 and the load volatility information by the maximum endpoint value of the load reduction parameter space to adjust the load reduction parameter space, performs load reduction optimization, obtains optimal load reduction parameters, and performs load decomposition on the power terminal.

2. The power terminal overload decomposition system with a self-repairing mechanism according to claim 1 is characterized in that: In the power terminal, load monitoring is performed. When overload occurs, continuous detection is performed to obtain the overload parameter sequence, including: In the power terminal, load monitoring is performed to determine whether the load is greater than the load threshold; If yes, then overload parameter recording is performed, and the overload parameters within the latest N moments are taken as the overload parameter sequence; if no, then load monitoring is continued, where N is an integer greater than 5.

3. The power terminal overload decomposition system with a self-repairing mechanism according to claim 1 is characterized in that: According to the overload parameter sequence, load scale analysis and load scale volatility analysis are performed to obtain load scale information and load volatility information, including: According to the overload parameter sequence, an average overload parameter is calculated to obtain load scale information; The load fluctuation information is calculated based on the overload parameter sequence and the load scale information.

4. The power terminal overload decomposition system with a self-repairing mechanism according to claim 3 is characterized in that: According to the overload parameter sequence and load scale information, load fluctuation information is calculated and obtained, including: Randomly selecting a first overload parameter in the overload parameter sequence, calculating a difference between the first overload parameter and the load scale information, and calculating a ratio of the difference to the load scale information as a first load fluctuation amplitude; Continue to randomly select and calculate the load fluctuation amplitudes of several overload parameters, calculate the mean, and obtain the load fluctuation information.

5. The power terminal overload decomposition system with a self-repairing mechanism according to claim 1 is characterized in that: According to the overload parameter sequence, the overload component damage probability analysis in the power terminal is performed to obtain basic component damage probability and damage probability volatility information, including: pre-training an overload component damage predictor of the power terminal; Using the overload component damage predictor, predicting component damage probabilities for multiple overload parameters in the overload parameter sequence to obtain multiple component damage probabilities; Calculating the average of the plurality of component damage probabilities to obtain a basic component damage probability; Randomly select a number of component damage probabilities from the multiple component damage probabilities, calculate the damage probability fluctuation amplitudes with the basic component damage probability respectively, and calculate the mean to obtain damage probability fluctuation information; The sum of 1 and the damage probability fluctuation information is multiplied by the basic component damage probability to perform a correction calculation to obtain the component damage probability.

6. The power terminal overload decomposition system with a self-repairing mechanism according to claim 5 is characterized in that: Pre-training the overload component damage predictor of the power terminal, comprising: According to the overload operation record data of the same model of power terminals, a sample overload parameter set is collected, and the proportion of power terminal component damage under different sample overload parameter operations is collected as a sample component damage probability set; The sample overload parameter set and the sample component damage probability set are used as supervised training data and verification data to train the overload component damage predictor until convergence.

7. The power terminal overload decomposition system with a self-repairing mechanism according to claim 6 is characterized in that: When the component damage probability is greater than a damage probability threshold, a load reduction parameter space is established according to the load scale information, the load reduction parameter space is adjusted according to the load volatility information, and load reduction optimization is performed, including: Determine whether the damage probability of the component is greater than a damage probability threshold, and if not, do not perform load decomposition; If yes, the load scale information is used as the maximum endpoint value of the load reduction parameter space to construct the load reduction parameter space; The maximum endpoint value of the load reduction parameter space is multiplied by the sum of 1 and the load fluctuation information to make an adjustment to obtain an adjusted load reduction parameter space; In the adjusted load reduction parameter space, a first load reduction parameter is randomly generated, and a component damage probability analysis after the load reduction is performed to obtain a first component damage probability, and a first decomposition fitness is calculated; Continue to optimize the load reduction parameters in the load reduction parameter space until convergence, output the load reduction parameters with the maximum decomposition fitness, obtain the optimal load reduction parameters, and perform load decomposition for load reduction on the power terminal.

8. The power terminal overload decomposition system with a self-repairing mechanism according to claim 7 is characterized in that: Perform component damage probability analysis after load reduction to obtain the first component damage probability and calculate the first decomposition fitness, including: Calculating the difference between the load scale information and the first load reduction parameter to obtain the first load parameter after the load reduction, inputting the difference into the overload component damage predictor, and analyzing and predicting to obtain the first component damage probability; According to the first component damage probability and the first load reduction parameter, the first decomposition fitness is calculated as follows: ； Among them, PF is the decomposition fitness, and is the weight, K is the component damage probability, is the load reduction parameter, and PA is the load parameter after load reduction.

Images