Wind power variable pitch super capacitor life health analysis method based on big data model

Abstract

The invention discloses a wind power variable pitch super capacitor life health analysis method based on a big data model, and relates to the field of maintenance of wind power generation equipment, and the method comprises the following operation steps: S1, collecting multi-source data; s2, data depth preprocessing; s3, carrying out feature engineering; s4, constructing and training a big data model; and S5, life health assessment and prediction. According to the big data model-based life health analysis method for the wind power variable-pitch super capacitor, the life health condition of the wind power variable-pitch super capacitor can be analyzed and predicted more comprehensively and accurately in real time, a powerful guarantee is provided for operation of wind power equipment, and by optimizing a sensor network, the service life of the wind power variable-pitch super capacitor is effectively improved. According to the method, the comprehensiveness and representativeness of data acquisition can be improved, meanwhile, comments of data acquisition can be dynamically adjusted, so that the utilization rate of energy consumption can be improved, and meanwhile, the accuracy of life analysis can be improved based on deep preprocessing of the data.

Description

Technical Field

[0001] The present invention relates to the field of maintenance of wind power generation equipment, and in particular to a life health analysis method for wind power variable pitch supercapacitors based on a big data model. Background Art

[0002] Wind turbine pitch supercapacitor is one of the key components of the wind turbine pitch system. In the wind turbine pitch system, when the wind speed changes, the angle of the blades needs to be adjusted to control the power output and speed of the wind turbine to ensure its stable operation within a safe range. Supercapacitors play a role in energy storage and rapid release in this process.

[0003] When analyzing the life health of wind turbine variable pitch supercapacitors, multi-source data cannot be obtained. At the same time, data may be missing during transmission, which will affect subsequent analysis and cannot comprehensively reflect the performance changes of supercapacitors from multiple angles. As a result, the accuracy of the evaluation cannot be guaranteed when analyzing and predicting the life health of supercapacitors.

[0004] Therefore, it is necessary to propose a life health analysis method for wind turbine pitch supercapacitors based on big data models to solve the above problems. Summary of the invention

[0005] The main purpose of the present invention is to provide a wind turbine variable pitch supercapacitor life health analysis method based on a big data model, which can effectively solve the problems in the background technology.

[0006] To achieve the above object, the technical solution adopted by the present invention is: The life health analysis method of wind turbine variable pitch supercapacitor based on big data model includes the following steps: S1: Multi-source data collection, through temperature, voltage, current sensors, humidity sensors, pressure sensors, the surface impact and potential impact of supercapacitors, data collection, using wireless sensor network technology to achieve real-time transmission and centralized collection of sensor data; S2: Data deep preprocessing, correcting the collected voltage and current data based on the model; S3: Feature engineering, extracting the waveform characteristics of the supercapacitor charging and discharging current and voltage, and constructing working condition adaptive features based on working condition information such as wind speed, wind direction, and blade angle; S4: Construction and training of big data models, preprocessing data based on the constructed hybrid models, including support vector machines, random forests and LSTM models, and improving the performance and stability of the models through training; S5: Life health assessment and prediction, based on the health status assessment of numerical indicators, image and sound modality information, a comprehensive assessment of life health is performed, and the remaining life of the supercapacitor is dynamically predicted based on the dynamic life prediction model.

[0007] Preferably, in S1, the collection of multi-source data includes a data collection module, a sensor network optimization module, and a data collection frequency dynamic adjustment module, and the collection of multi-source data is based on temperature, voltage, current sensors, humidity sensors, and pressure sensors to collect real-time data on the temperature, humidity, voltage, current, and pressure of the supercapacitor; The sensor network optimization module distributes temperature, voltage, current sensors, humidity sensors, and pressure sensors based on a grid layout to ensure that accurate data can be obtained at each key part of the supercapacitor, thereby improving the comprehensiveness and representativeness of the data; The data acquisition frequency dynamic adjustment module dynamically adjusts the data acquisition frequency according to the operating status of the variable pitch system. During the normal and stable operation stage of the variable pitch system, the acquisition frequency is reduced to once per second to reduce the pressure of data storage and transmission; under critical working conditions where the wind speed suddenly changes and the pitch action is frequent, the acquisition frequency is automatically increased to 10 times per second to ensure that the detailed performance changes of the supercapacitor under complex working conditions are captured, and the normal frequency is restored after continuous acquisition for 3 minutes.

[0008] Preferably, S2 includes the following steps: S201: Based on the data correction of the physical model, the collected voltage and current data are corrected by using the electrochemical reaction principle and equivalent circuit model of the supercapacitor, including establishing a polarized capacitor model to compensate the measured capacitor voltage to accurately reflect the actual charge state of the capacitor; according to the heat conduction model, combined with the temperature sensor data and the heat dissipation structure parameters of the pitch control system, the internal temperature of the supercapacitor is estimated and corrected to obtain temperature data closer to the actual situation and improve the accuracy of subsequent life analysis; S202: Noise filtering and feature enhancement. Use signal processing technology to filter the data for noise. For high-frequency noise interference in the current signal, remove it through the wavelet threshold denoising method to make the current curve smoother for subsequent feature extraction and analysis. Use data enhancement technology to transform and expand the original data, including translating, scaling and adding small random noise operations to the voltage data, to generate more training data samples and improve the robustness and generalization ability of the model.

[0009] Preferably, S3 includes the following steps: S301: Based on the feature extraction of time series, the statistical characteristics of supercapacitors at different time scales are calculated, including the average charge and discharge current, voltage standard deviation, temperature range, etc. at the hourly, daily and weekly levels, which are used to reflect the performance stability and change trend of supercapacitors during long-term operation; waveform characteristics of charge and discharge current and voltage are extracted, including the rising edge slope, falling edge slope, peak time, and pulse width, which are used to reveal the electrochemical reaction process inside the supercapacitor and the performance changes of the electrode materials; S302: Construction of working condition adaptive features. Based on the working condition information of wind speed, wind direction and blade angle, working condition adaptive features are constructed, including definition of equivalent wind speed impact times feature, that is, according to the wind speed change and blade angle adjustment, the equivalent charge and discharge times of the supercapacitor under different wind speed impacts are counted to accurately reflect the life consumption of the supercapacitor under actual working conditions. Combined with the operating mode of the variable pitch system, operating mode related features are created, including calculation of the instantaneous energy release rate and voltage drop amplitude of the supercapacitor in the emergency braking mode, which are used to evaluate the health of the supercapacitor under extreme working conditions.

[0010] Preferably, the step S4 includes the following steps: S401: Integrate multiple machine learning models to build a hybrid model, integrating support vector machine, random forest and LSTM models. First, use support vector machine model and random forest model to perform preliminary classification and regression analysis on preprocessed feature data to extract linear and nonlinear features in the data. Use these features as input to the LSTM model, mine time series data based on LSTM, and mine long-term dependencies and dynamic change rules in the data. Use the AdaBoost algorithm to perform weighted combination of multiple support vector machine models to improve the overall performance and stability of the model. During the training process, dynamically adjust the weight of each model according to its prediction error, so that the model with better prediction performance has a greater say in the final decision. S402: Model training optimization based on reinforcement learning, introducing the Q-Learning algorithm to optimize the model training process, using the model's prediction accuracy and stability as reward signals, and automatically finding the optimal training strategy by training and exploring the model on different hyperparameter configurations and data subsets.

[0011] Preferably, the step S5 includes the following steps: S501: Multimodal health status assessment, through health status assessment based on numerical indicators, image and sound modal information for comprehensive assessment, using infrared thermal imaging technology to obtain the surface temperature distribution image of the supercapacitor, and using image analysis algorithm to detect whether there is a local overheating area, which is used to evaluate the short circuit and poor contact problems inside the supercapacitor. At the same time, the sound signal of the supercapacitor during operation is collected, and the audio feature extraction and classification technology is used to determine whether there is abnormal noise, including but not limited to the buzzing sound caused by electrolyte leakage and the collision sound caused by loose electrodes. Combined with the evaluation results of the physical model and the big data model, the Dempster-Shafer evidence theory is used for information fusion, and the theoretical life value calculated by the physical model and the life value predicted by the big data model are used as different evidence sources. The synthesis rules of the evidence theory are used to obtain the supercapacitor life health assessment results; S502: Dynamic life prediction and confidence interval estimation. Based on the dynamic changes in the operating environment and working conditions of the wind turbine variable pitch system, a dynamic life prediction model is established. The Kalman filtering method is used to continuously update the model parameters and state estimates based on real-time monitoring data, thereby realizing dynamic prediction of the remaining life of the supercapacitor, including real-time adjustment of the predicted value of the remaining life as the wind speed changes and the supercapacitor performance decays. At the same time, the Bayesian statistical method is used to estimate the confidence interval of the life prediction. Through statistical analysis of historical data and uncertainty modeling of the model, the confidence level of the remaining life prediction value is given, providing more comprehensive information for operation and maintenance decisions, and issuing early warnings based on the actual remaining life prediction value for subsequent maintenance of the supercapacitor.

[0012] Compared with the prior art, the present invention provides a life health analysis method for wind turbine pitch supercapacitors based on a big data model, which has the following beneficial effects: The wind turbine variable pitch supercapacitor life health analysis method based on the big data model can analyze and predict the life health status of wind turbine variable pitch supercapacitors in a more comprehensive, accurate and real-time manner, providing strong guarantees for the operation of wind power equipment. By optimizing the sensor network, the comprehensiveness and representativeness of data collection can be improved, and the comments on data collection can be dynamically adjusted to improve the utilization rate of energy consumption. At the same time, it can be based on deep preprocessing of data to improve the accuracy of life analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0014] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0015] Embodiment 1: like Figure 1 As shown in the figure, the life health analysis method of wind turbine variable pitch supercapacitor based on big data model includes the following steps: S1: Multi-source data collection, through temperature, voltage, current sensors, humidity sensors, pressure sensors, the surface impact and potential impact of supercapacitors, data collection, using wireless sensor network technology to achieve real-time transmission and centralized collection of sensor data; S2: Data deep preprocessing, correcting the collected voltage and current data based on the model; S3: Feature engineering, extracting the waveform characteristics of the supercapacitor charging and discharging current and voltage, and constructing working condition adaptive features based on working condition information such as wind speed, wind direction, and blade angle; S4: Construction and training of big data models, preprocessing data based on the constructed hybrid models, including support vector machines, random forests and LSTM models, and improving the performance and stability of the models through training; S5: Life health assessment and prediction, based on the health status assessment of numerical indicators, image and sound modality information, a comprehensive assessment of life health is performed, and the remaining life of the supercapacitor is dynamically predicted based on the dynamic life prediction model.

[0016] Embodiment 2: A life health analysis method for wind turbine variable pitch supercapacitors based on a big data model. The collection of multi-source data uses temperature, voltage, current sensors, humidity sensors, and pressure sensors to collect data on the surface impact and potential impact of supercapacitors. The wireless sensor network technology is used to achieve real-time transmission and centralized collection of sensor data. The collection of multi-source data includes data collection module, sensor network optimization module, and data collection frequency dynamic adjustment module. The collection of multi-source data is based on temperature, voltage, current sensors, humidity sensors, and pressure sensors to collect real-time data on the temperature, humidity, voltage, current, and pressure of the supercapacitor; The sensor network optimization module distributes temperature, voltage, current sensors, humidity sensors, and pressure sensors based on a grid layout to ensure accurate data can be obtained at each key part of the supercapacitor, improving the comprehensiveness and representativeness of the data; The data acquisition frequency dynamic adjustment module dynamically adjusts the data acquisition frequency according to the operating status of the variable pitch system. During the normal and stable operation stage of the variable pitch system, the acquisition frequency is reduced to once per second to reduce the pressure of data storage and transmission. Under critical working conditions such as sudden wind speed changes and frequent pitch changes, the acquisition frequency is automatically increased to 10 times per second to ensure that the detailed performance changes of the supercapacitor under complex working conditions are captured, and the data is restored to the normal frequency after 3 minutes of continuous acquisition.

[0017] Embodiment three: The life health analysis method of wind turbine variable pitch supercapacitor based on big data model, deep data preprocessing, and correction of collected voltage and current data based on the model include the following steps: S201: Based on the data correction of the physical model, the collected voltage and current data are corrected by using the electrochemical reaction principle and equivalent circuit model of the supercapacitor, including establishing a polarized capacitor model to compensate the measured capacitor voltage to accurately reflect the actual charge state of the capacitor; according to the heat conduction model, combined with the temperature sensor data and the heat dissipation structure parameters of the pitch control system, the internal temperature of the supercapacitor is estimated and corrected to obtain temperature data closer to the actual situation and improve the accuracy of subsequent life analysis; S202: Noise filtering and feature enhancement. Use signal processing technology to filter the data for noise. For high-frequency noise interference in the current signal, remove it through the wavelet threshold denoising method to make the current curve smoother for subsequent feature extraction and analysis. Use data enhancement technology to transform and expand the original data, including translating, scaling and adding small random noise operations to the voltage data, to generate more training data samples and improve the robustness and generalization ability of the model.

[0018] Embodiment 4: The life health analysis method of wind turbine variable pitch supercapacitor based on big data model, feature engineering, extracting the waveform characteristics of supercapacitor charging and discharging current and voltage, and constructing working condition adaptive features according to working condition information such as wind speed, wind direction and blade angle, including the following steps: S301: Based on the feature extraction of time series, the statistical characteristics of supercapacitors at different time scales are calculated, including the average charge and discharge current, voltage standard deviation, temperature range, etc. at the hourly, daily and weekly levels, which are used to reflect the performance stability and change trend of supercapacitors during long-term operation; waveform characteristics of charge and discharge current and voltage are extracted, including the rising edge slope, falling edge slope, peak time, and pulse width, which are used to reveal the electrochemical reaction process inside the supercapacitor and the performance changes of the electrode materials; S302: Construction of working condition adaptive features. According to the working condition information of wind speed, wind direction and blade angle, working condition adaptive features are constructed, including definition of equivalent wind speed impact times feature, that is, according to the wind speed change and blade angle adjustment, the equivalent charge and discharge times of the supercapacitor under different wind speed impacts are counted, which is used to accurately reflect the life consumption of the supercapacitor under actual working conditions. Combined with the operating mode of the variable pitch system, operating mode related features are created, including calculation of the instantaneous energy release rate and voltage drop amplitude of the supercapacitor in the emergency braking mode, which is used to evaluate the health of the supercapacitor under extreme working conditions.

[0019] Embodiment five: A life health analysis method for wind turbine variable pitch supercapacitors based on a big data model, the construction and training of a big data model, preprocessing of data based on the constructed hybrid model, including support vector machine, random forest and LSTM models, and improving the performance and stability of the model through training, including the following steps: S401: Integrate multiple machine learning models to build a hybrid model, integrating support vector machine, random forest and LSTM models. First, use support vector machine model and random forest model to perform preliminary classification and regression analysis on preprocessed feature data to extract linear and nonlinear features in the data. Use these features as input to the LSTM model, mine time series data based on LSTM, and mine long-term dependencies and dynamic change rules in the data. Use the AdaBoost algorithm to perform weighted combination of multiple support vector machine models to improve the overall performance and stability of the model. During the training process, dynamically adjust the weight of each model according to its prediction error, so that the model with better prediction performance has a greater say in the final decision. S402: Model training optimization based on reinforcement learning, introducing the Q-Learning algorithm to optimize the model training process, using the model's prediction accuracy and stability as reward signals, and automatically finding the optimal training strategy by training and exploring the model on different hyperparameter configurations and data subsets.

[0020] Embodiment six: The life health analysis method of wind turbine variable pitch supercapacitor based on big data model, life health assessment and prediction, health status assessment based on numerical indicators, image and sound modal information, and dynamic prediction of the remaining life of supercapacitor based on dynamic life prediction model include the following steps: S501: Multimodal health status assessment, through health status assessment based on numerical indicators, image and sound modal information for comprehensive assessment, using infrared thermal imaging technology to obtain the surface temperature distribution image of the supercapacitor, and using image analysis algorithm to detect whether there is a local overheating area, which is used to evaluate the short circuit and poor contact problems inside the supercapacitor. At the same time, the sound signal of the supercapacitor during operation is collected, and the audio feature extraction and classification technology is used to determine whether there is abnormal noise, including but not limited to the buzzing sound caused by electrolyte leakage and the collision sound caused by loose electrodes. Combined with the evaluation results of the physical model and the big data model, the Dempster-Shafer evidence theory is used for information fusion, and the theoretical life value calculated by the physical model and the life value predicted by the big data model are used as different evidence sources. The synthesis rules of the evidence theory are used to obtain the supercapacitor life health assessment results; S502: Dynamic life prediction and confidence interval estimation. Based on the dynamic changes in the operating environment and working conditions of the wind turbine variable pitch system, a dynamic life prediction model is established. The Kalman filtering method is used to continuously update the model parameters and state estimates based on real-time monitoring data, thereby realizing dynamic prediction of the remaining life of the supercapacitor, including real-time adjustment of the predicted value of the remaining life as the wind speed changes and the supercapacitor performance decays. At the same time, the Bayesian statistical method is used to estimate the confidence interval of the life prediction. Through statistical analysis of historical data and uncertainty modeling of the model, the confidence level of the remaining life prediction value is given, providing more comprehensive information for operation and maintenance decisions, and issuing early warnings based on the actual remaining life prediction value for subsequent maintenance of the supercapacitor.

[0021] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of 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 to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A life health analysis method for wind turbine variable pitch supercapacitors based on a big data model, characterized by: The steps include: S1: Multi-source data collection, through temperature, voltage, current sensors, humidity sensors, pressure sensors, the surface impact and potential impact of supercapacitors, data collection, using wireless sensor network technology to achieve real-time transmission and centralized collection of sensor data; S2: Data deep preprocessing, correcting the collected voltage and current data based on the model; S3: Feature engineering, extracting the waveform characteristics of the supercapacitor charging and discharging current and voltage, and constructing working condition adaptive features based on working condition information such as wind speed, wind direction, and blade angle; S4: Construction and training of big data models, preprocessing data based on the constructed hybrid models, including support vector machines, random forests and LSTM models, and improving the performance and stability of the models through training; S5: Life health assessment and prediction, based on the health status assessment of numerical indicators, image and sound modality information, a comprehensive assessment of life health is performed, and the remaining life of the supercapacitor is dynamically predicted based on the dynamic life prediction model.

2. The wind turbine variable pitch supercapacitor life health analysis method based on big data model according to claim 1 is characterized by: In S1, the collection of multi-source data includes a data collection module, a sensor network optimization module, and a data collection frequency dynamic adjustment module. The multi-source data collection is based on temperature, voltage, current sensors, humidity sensors, and pressure sensors to collect real-time data on the temperature, humidity, voltage, current, and pressure of the supercapacitor; The sensor network optimization module distributes temperature, voltage, current sensors, humidity sensors, and pressure sensors based on a grid layout to ensure that accurate data can be obtained at each key part of the supercapacitor, thereby improving the comprehensiveness and representativeness of the data; The data acquisition frequency dynamic adjustment module dynamically adjusts the data acquisition frequency according to the operating status of the variable pitch system. During the normal and stable operation stage of the variable pitch system, the acquisition frequency is reduced to once per second to reduce the pressure of data storage and transmission; under critical working conditions where the wind speed suddenly changes and the pitch action is frequent, the acquisition frequency is automatically increased to 10 times per second to ensure that the detailed performance changes of the supercapacitor under complex working conditions are captured, and the normal frequency is restored after continuous acquisition for 3 minutes.

3. The wind turbine variable pitch supercapacitor life health analysis method based on big data model according to claim 1 is characterized by: The S2 includes the following steps: S201: Data correction based on the physical model, using the electrochemical reaction principle and equivalent circuit model of the supercapacitor, corrects the collected voltage and current data, including establishing a polarization capacitor model and compensating the measured capacitor voltage to accurately reflect the actual charge state of the capacitor; According to the heat conduction model, combined with the temperature sensor data and the heat dissipation structure parameters of the pitch control system, the internal temperature of the supercapacitor is estimated and corrected to obtain temperature data closer to the actual situation and improve the accuracy of subsequent life analysis; S202: Noise filtering and feature enhancement. Use signal processing technology to filter the data for noise. For high-frequency noise interference in the current signal, remove it through the wavelet threshold denoising method to make the current curve smoother for subsequent feature extraction and analysis. Use data enhancement technology to transform and expand the original data, including translating, scaling and adding small random noise operations to the voltage data, to generate more training data samples and improve the robustness and generalization ability of the model.

4. The wind turbine variable pitch supercapacitor life health analysis method based on big data model according to claim 1 is characterized by: The S3 includes the following steps: S301: Based on the feature extraction of time series, the statistical characteristics of supercapacitors at different time scales are calculated, including the average charge and discharge current, voltage standard deviation, temperature range, etc. at the hourly, daily and weekly levels, which are used to reflect the performance stability and change trend of supercapacitors during long-term operation; waveform characteristics of charge and discharge current and voltage are extracted, including the rising edge slope, falling edge slope, peak time, and pulse width, which are used to reveal the electrochemical reaction process inside the supercapacitor and the performance changes of the electrode materials; S302: Construction of working condition adaptive features. According to the working condition information of wind speed, wind direction and blade angle, working condition adaptive features are constructed, including definition of equivalent wind speed impact times feature, that is, according to the wind speed change and blade angle adjustment, the equivalent charge and discharge times of the supercapacitor under different wind speed impacts are counted, which is used to accurately reflect the life consumption of the supercapacitor under actual working conditions. Combined with the operating mode of the variable pitch system, operating mode related features are created, including calculation of the instantaneous energy release rate and voltage drop amplitude of the supercapacitor in the emergency braking mode, which is used to evaluate the health of the supercapacitor under extreme working conditions.

5. The wind turbine variable pitch supercapacitor life health analysis method based on big data model according to claim 1 is characterized by: The S4 includes the following steps: S401: Integrate multiple machine learning models to build a hybrid model, integrating support vector machine, random forest and LSTM models. First, use the support vector machine model and random forest model to perform preliminary classification and regression analysis on the preprocessed feature data to extract linear and nonlinear features in the data; These features are used as the input of the LSTM model. The time series data is mined based on LSTM to mine the long-term dependencies and dynamic change rules in the data. Multiple support vector machine models are weighted and combined through the AdaBoost algorithm to improve the overall performance and stability of the model. During the training process, the weight of each model is dynamically adjusted according to its prediction error, so that the model with better prediction performance has a greater say in the final decision. S402: Model training optimization based on reinforcement learning, introducing the Q-Learning algorithm to optimize the model training process, using the model's prediction accuracy and stability as reward signals, and automatically finding the optimal training strategy by training and exploring the model on different hyperparameter configurations and data subsets.

6. The wind turbine variable pitch supercapacitor life health analysis method based on big data model according to claim 1 is characterized by: The S5 includes the following steps: S501: Multimodal health status assessment, through health status assessment based on numerical indicators, image and sound modal information for comprehensive assessment, using infrared thermal imaging technology to obtain the surface temperature distribution image of the supercapacitor, and using image analysis algorithm to detect whether there is a local overheating area, which is used to evaluate the short circuit and poor contact problems inside the supercapacitor. At the same time, the sound signal of the supercapacitor during operation is collected, and the audio feature extraction and classification technology is used to determine whether there is abnormal noise, including but not limited to the buzzing sound caused by electrolyte leakage and the collision sound caused by loose electrodes. Combined with the evaluation results of the physical model and the big data model, the Dempster-Shafer evidence theory is used for information fusion, and the theoretical life value calculated by the physical model and the life value predicted by the big data model are used as different evidence sources. The synthesis rules of the evidence theory are used to obtain the supercapacitor life health assessment results; S502: Dynamic life prediction and confidence interval estimation. Based on the dynamic changes in the operating environment and working conditions of the wind turbine variable pitch system, a dynamic life prediction model is established. The Kalman filtering method is used to continuously update the model parameters and state estimates based on real-time monitoring data, thereby realizing dynamic prediction of the remaining life of the supercapacitor, including real-time adjustment of the predicted value of the remaining life as the wind speed changes and the supercapacitor performance decays. At the same time, the Bayesian statistical method is used to estimate the confidence interval of the life prediction. Through statistical analysis of historical data and uncertainty modeling of the model, the confidence level of the remaining life prediction value is given, providing more comprehensive information for operation and maintenance decisions, and issuing early warnings based on the actual remaining life prediction value for subsequent maintenance of the supercapacitor.

Images