New energy charging simulation optimization method and device based on charging and discharging test platform

Abstract

The invention relates to the field of power simulation optimization, in particular to a new energy charging simulation optimization method and device based on a charging and discharging test platform. The method comprises the following steps of performing new energy charging pile charging simulation based on a charging and discharging test platform, performing adjustable power time period analysis, and generating a plurality of adjustable power windows; performing timing sequence energy flow trend change mining and dynamic power distribution fitting according to the plurality of adjustable power windows, and constructing a dynamic power path distribution map; performing minimum power drift point detection according to the dynamic power path distribution map, and performing power fluctuation compensation inversion to obtain power fluctuation compensation configuration parameters; and performing local dynamic power regulation according to the power fluctuation compensation configuration parameters, and constructing a local window dynamic power regulation strategy. Through dynamic charging simulation, the power stability of new energy charging is improved, and the power response balance is optimized.

Description

Technical Field

[0001] The present invention relates to the field of power simulation optimization, and in particular to a new energy charging simulation optimization method and device based on a charge and discharge test platform. Background Art

[0002] With the rapid development of the new energy vehicle industry, charging technology, as a key component supporting the efficient operation of electric transportation systems, has gradually become a hot topic in research and engineering practice. To meet the diverse charging needs of electric vehicles and the flexible load regulation requirements of the power grid, charging pile systems are evolving towards high power, high efficiency, and intelligent technology. In this process, how to accurately simulate and dynamically optimize charging power has become a core issue for improving energy management efficiency and ensuring stable system operation.

[0003] Currently, new energy charging systems generally face technical challenges such as complex load environments, significant power fluctuations, and variable operating conditions. During the charging process, the power output of charging piles is often affected by multiple factors, including grid disturbances, battery status, and charging strategies. This results in a significant dynamic change in the power curve. These changes not only affect charging efficiency but can also negatively impact the power grid, such as power surges and load imbalances. This further exacerbates operational uncertainty and safety risks in new energy charging systems.

[0004] Traditional charging power control and simulation methods mostly rely on static test data or rule-based parameter adjustments and lack dynamic perception and adaptive optimization capabilities. Existing methods struggle to fully identify and optimize complex power change paths, especially under conditions of multi-scenario interaction and multiple load disturbances. Furthermore, these methods generally lack high-frequency, multi-granularity power behavior monitoring capabilities, making it impossible to effectively extract key fluctuation characteristics and control nodes during the charging process, severely limiting simulation accuracy and system responsiveness. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention proposes a new energy charging simulation optimization method and device based on a charge and discharge test platform to solve at least one of the above technical problems.

[0006] To achieve the above objectives, the present invention provides a new energy charging simulation optimization method based on a charge and discharge test platform, comprising the following steps: Step S1: performing a new energy charging pile charging simulation based on the charging and discharging test platform, and performing an adjustable power time period analysis to generate multiple adjustable power windows; Step S2: mining the trend of time-series energy flow and fitting the dynamic power distribution according to the multiple adjustable power windows to construct a dynamic power path distribution map; Step S3: performing minimum power drift point detection according to the dynamic power path distribution map, and performing power fluctuation compensation inversion to obtain power fluctuation compensation configuration parameters; Step S4: performing local dynamic power regulation according to the power fluctuation compensation configuration parameters and constructing a local window dynamic power control strategy; Step S5: performing charging simulation optimization according to the local window dynamic power control strategy, and performing power control feedback evaluation to generate power control feedback information; Step S6: Perform global power control optimization and adaptive power feedback evolution according to the power control feedback information, and build an adaptive power optimization model.

[0007] This invention uses a charging and discharging test platform to simulate the actual charging process of a new energy charging pile under different load conditions and with different vehicle battery capacities and types. It collects real-time multi-dimensional data, including voltage, current, and power. It then uses a sliding window algorithm or fluctuation analysis method to perform time-series analysis on the power curve. Based on statistical feature extraction (mean, standard deviation, coefficient of variation, etc.), it identifies stable and adjustable power regions and generates multiple power-adjustable time periods (adjustable power windows). This method deeply reveals the temporal nature and fluctuation trends of energy flow within these windows, improving control accuracy. By constructing a "dynamic power path distribution map," it visualizes power changes, facilitating the identification of key nodes for local or global regulation. This improves the predictive capabilities of power management and provides a basis for decision-making on power regulation tailored to local and time conditions. Based on the adjustable power windows, data mining techniques are used to identify trends in the time-series energy flow (the energy transfer path from the grid to the vehicle) during the charging process. Key time nodes, such as power peaks, valleys, and sudden changes, can be identified, providing a predictive basis for subsequent control strategies. By fitting these time series data (e.g., using polynomial fitting, Bayesian filtering, spline functions, etc.), a high-resolution "dynamic power path distribution map" can be constructed. Identifying the minimum point of power drift can promptly locate critical points of system stability, preventing system anomalies caused by sudden power changes. Fluctuation compensation inversion dynamically calculates adaptable compensation parameters, improving responsiveness to power fluctuations, enhancing system robustness under complex or non-ideal load conditions, and improving the overall smoothness of the charging process. Based on the compensation configuration parameters, the previously identified power windows are adjusted in real time at the local level. The system uses a control algorithm to fine-tune charging power within each power window, ensuring dynamic adaptation to changes in user charging behavior, grid demand, or load conditions while maintaining system stability. The constructed local window dynamic power control strategy features a short timescale, high granularity, and a narrow control range. It enables rapid response and precise control of local power fluctuations, significantly improving system control efficiency at the microscopic timescale. A second simulation test of actual charging scenarios was conducted in a simulation environment to verify the strategy's effectiveness, adaptability, and mitigation of power fluctuations. During the test, real-time system operating status data under different operating conditions is collected and multi-dimensional feedback evaluation is performed, including key indicators such as response delay, compensation accuracy, user charging efficiency, and grid load balance. This information is then used to generate system-level power control feedback information. By integrating the feedback evaluation results of multiple local strategies, the power control logic is globally optimized from the overall architecture level, considering multiple objective functions such as system energy efficiency, grid security, and user satisfaction. By introducing reinforcement learning, adaptive control, or neural network optimization models, the feedback information is iteratively trained over multiple rounds to construct a self-evolving adaptive power optimization model.This model is adaptable to diverse application scenarios and can dynamically update control strategies during system operation, enhancing the intelligence and self-learning capabilities of the charging control system, thereby achieving true "intelligent charging control." Furthermore, through continuous learning and parameter self-adjustment, the model can quickly adapt to new loads, varying electricity pricing mechanisms, or policy environments, significantly improving the system's long-term operational performance.

[0008] In this specification, a new energy charging simulation optimization device based on a charge and discharge test platform is provided, which is used to execute the new energy charging simulation optimization method based on the charge and discharge test platform as described above, including: A charging simulation module is used to simulate the charging of a new energy charging pile based on the charging and discharging test platform, analyze the adjustable power time period, and generate multiple adjustable power windows; A power path distribution module is used to mine the trend of time-series energy flow and fit the dynamic power distribution according to the multiple adjustable power windows, and to construct a dynamic power path distribution map; a fluctuation compensation module, configured to detect a minimum power drift point according to the dynamic power path distribution map and perform power fluctuation compensation inversion to obtain power fluctuation compensation configuration parameters; The local adjustment module is used to perform local dynamic power adjustment based on the power fluctuation compensation configuration parameters and build a local window dynamic power control strategy; The control feedback module is used to perform charging simulation optimization based on the local window dynamic power control strategy, and perform power control feedback evaluation to generate power control feedback information; The global control module is used to optimize global power control and adaptive power feedback evolution based on power control feedback information, and build an adaptive power optimization model The present invention analyzes the stability, volatility and other characteristics of the power curve during the charging process, identifies multiple adjustable power time periods, generates corresponding power adjustment windows, and provides boundary conditions and decision support for subsequent optimization operations. Avoid over-abstract modeling, achieve realistic simulation of the actual operating state, and improve the applicability and effectiveness of the power control strategy. Deeply explore the power change trajectory within each adjustable time window, capture detailed information such as the direction, amplitude, and frequency of the energy flow, and form time series trend data. Through mathematical modeling and fitting analysis, a "dynamic power path distribution map" is established to vividly display the changing process of power flow, providing a visual and quantifiable basis for refined control. Improve the understanding of the power fluctuation mechanism during the charging process and lay a data foundation for subsequent predictive control and adaptive adjustment. Accurately identify the critical point where power drift occurs during the charging process, which serves as the "grasp" of the adjustment strategy to achieve precise control. Introduce an inversion analysis mechanism to mathematically model the causes of fluctuations and back-derive parameters, and derive scientific and reasonable power fluctuation compensation configuration parameters. This system proactively suppresses potential large fluctuations, ensuring stable power output and improving charging equipment operational stability and power control sophistication. It flexibly implements power regulation within specific time windows and regions, avoiding resource waste and response delays associated with unified global control. It fine-grainedly divides control strategies based on real-time status and compensation parameters, improving control response speed and scenario adaptability. This helps reduce local load pressure, balance charging peak loads, and enhance the flexibility and practicality of the power control system. Backtesting and data feedback analysis are performed on local power control strategies, forming a closed-loop evaluation mechanism for control effectiveness. Dynamically collects information on the system's response to control strategies, such as the degree of fluctuation suppression, efficiency improvement, and anomaly response capabilities. This enables continuous iterative optimization of control strategies, providing decision data and strategy improvement directions for subsequent model upgrades. By integrating historical and real-time feedback data to form a multi-dimensional decision-making basis, global power strategy optimization is implemented, achieving intelligent control at the macro level. By incorporating machine learning and adaptive adjustment algorithms, the system is equipped with the capabilities of policy learning, autonomous evolution, and self-correction. The resulting adaptive power optimization model is portable and scalable, applicable to various new energy scenarios, enabling the transition from local control to global coordination. The system has the ability to continuously evolve and can effectively respond to challenges such as equipment diversification, rapid grid response, and complex charging behavior in new energy scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 This is a schematic flow chart of the steps of a new energy charging simulation optimization method based on a charge and discharge test platform of the present invention; Figure 2 Detailed implementation flow chart of step S1; Figure 3 Detailed implementation flow chart of step S2; Figure 4 Schematic diagram of the detailed implementation steps of step S3. DETAILED DESCRIPTION

[0010] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0011] This application example provides a new energy charging simulation optimization method and device based on a charging and discharging test platform. The execution subjects of the new energy charging simulation optimization method and device based on the charging and discharging test platform include but are not limited to: mechanical equipment, data processing platforms, cloud server nodes, network upload devices, etc. equipped with the system, which can be regarded as general computing nodes of this application. The data processing platform includes but is not limited to: at least one of an audio and image management system, an information management system, and a cloud data management system.

[0012] See also Figures 1 to 4 The present invention provides a new energy charging simulation optimization method based on a charge and discharge test platform, and the new energy charging simulation optimization method based on the charge and discharge test platform includes the following steps: Step S1: performing a new energy charging pile charging simulation based on the charging and discharging test platform, and performing an adjustable power time period analysis to generate multiple adjustable power windows; Step S2: mining the trend of time-series energy flow and fitting the dynamic power distribution according to the multiple adjustable power windows to construct a dynamic power path distribution map; Step S3: performing minimum power drift point detection according to the dynamic power path distribution map, and performing power fluctuation compensation inversion to obtain power fluctuation compensation configuration parameters; Step S4: performing local dynamic power regulation according to the power fluctuation compensation configuration parameters and constructing a local window dynamic power control strategy; Step S5: performing charging simulation optimization according to the local window dynamic power control strategy, and performing power control feedback evaluation to generate power control feedback information; Step S6: Perform global power control optimization and adaptive power feedback evolution according to the power control feedback information, and build an adaptive power optimization model.

[0013] This invention uses a charging and discharging test platform to simulate the actual charging process of a new energy charging pile under different load conditions and with different vehicle battery capacities and types. It collects real-time multi-dimensional data, including voltage, current, and power. It then uses a sliding window algorithm or fluctuation analysis method to perform time-series analysis on the power curve. Based on statistical feature extraction (mean, standard deviation, coefficient of variation, etc.), it identifies stable and adjustable power regions and generates multiple power-adjustable time periods (adjustable power windows). This method deeply reveals the temporal nature and fluctuation trends of energy flow within these windows, improving control accuracy. By constructing a "dynamic power path distribution map," it visualizes power changes, facilitating the identification of key nodes for local or global regulation. This improves the predictive capabilities of power management and provides a basis for decision-making on power regulation tailored to local and time conditions. Based on the adjustable power windows, data mining techniques are used to identify trends in the time-series energy flow (the energy transfer path from the grid to the vehicle) during the charging process. Key time nodes, such as power peaks, valleys, and sudden changes, can be identified, providing a predictive basis for subsequent control strategies. By fitting these time series data (e.g., using polynomial fitting, Bayesian filtering, spline functions, etc.), a high-resolution "dynamic power path distribution map" can be constructed. Identifying the minimum point of power drift can promptly locate critical points of system stability, preventing system anomalies caused by sudden power changes. Fluctuation compensation inversion dynamically calculates adaptable compensation parameters, improving responsiveness to power fluctuations, enhancing system robustness under complex or non-ideal load conditions, and improving the overall smoothness of the charging process. Based on the compensation configuration parameters, the previously identified power windows are adjusted in real time at the local level. The system uses a control algorithm to fine-tune charging power within each power window, ensuring dynamic adaptation to changes in user charging behavior, grid demand, or load conditions while maintaining system stability. The constructed local window dynamic power control strategy features a short timescale, high granularity, and a narrow control range. It enables rapid response and precise control of local power fluctuations, significantly improving system control efficiency at the microscopic timescale. A second simulation test of actual charging scenarios was conducted in a simulation environment to verify the strategy's effectiveness, adaptability, and mitigation of power fluctuations. During the test, real-time system operating status data under different operating conditions is collected and multi-dimensional feedback evaluation is performed, including key indicators such as response delay, compensation accuracy, user charging efficiency, and grid load balance. This information is then used to generate system-level power control feedback information. By integrating the feedback evaluation results of multiple local strategies, the power control logic is globally optimized from the overall architecture level, considering multiple objective functions such as system energy efficiency, grid security, and user satisfaction. By introducing reinforcement learning, adaptive control, or neural network optimization models, the feedback information is iteratively trained over multiple rounds to construct a self-evolving adaptive power optimization model.This model is adaptable to diverse application scenarios and can dynamically update control strategies during system operation, enhancing the intelligence and self-learning capabilities of the charging control system, thereby achieving true "intelligent charging control." Furthermore, through continuous learning and parameter self-adjustment, the model can quickly adapt to new loads, varying electricity pricing mechanisms, or policy environments, significantly improving the system's long-term operational performance.

[0014] In the embodiment of the present invention, see Figure 1 , is a schematic flow chart of the steps of a new energy charging simulation optimization method based on a charge and discharge test platform of the present invention. In this example, the steps of the new energy charging simulation optimization method based on the charge and discharge test platform include: Step S1: performing a new energy charging pile charging simulation based on the charging and discharging test platform, and performing an adjustable power time period analysis to generate multiple adjustable power windows; In this example, a charging and discharging test platform was constructed to simulate the charging of a new energy charging pile. The test platform should include a charging pile, a load simulator, a data acquisition system, and monitoring instruments to ensure that it can simulate different charging conditions and grid load conditions. The maximum output power of the charging pile is set to 30 kW, and the minimum output power is set to 3 kW. The load simulator must be able to simulate grid load fluctuations, with a load variation range of ±10%. System integration is performed on the test platform to ensure effective communication between components and real-time data collection and monitoring capabilities. Before testing, system debugging is performed to ensure normal equipment operation. The data collection frequency is configured to sample once per second to obtain real-time data during the charging process. Charging simulation startup: Start the charging simulation program and set the charging time to 2 hours. Adjust the output power of the charging pile based on user needs and grid load conditions. During the simulation, monitor the charging pile's power output, voltage, current, and other parameters in real time. Set the initial charging power to 20 kW and gradually adjust it based on grid load conditions to ensure complete charging data collection during the simulation. During the charging process, real-time data is collected on various parameters, including charging power, charging time, and load fluctuations. Ensure data integrity and provide a foundation for subsequent analysis of adjustable power time periods. Record charging power data every minute to ensure a complete charging curve and load changes are captured at the end of the charging simulation. After the charging simulation, analyze the adjustable power time periods based on the collected data. First, define the criteria for the adjustable power time periods, typically periods with minimal power fluctuations or load fluctuations. Set the threshold for the adjustable power window to ±2 kW and examine which time periods throughout the charging process fall within this range. Process the charging power data to identify time periods that meet the adjustable criteria. This can be automated using programming tools or data analysis software. If the analysis results show that power remains stable between 18 kW and 20 kW for a certain period (e.g., from the 30th to the 40th minute), record this period as an adjustable power window. Record the identified adjustable power windows in a database, ensuring each window has a detailed time stamp and corresponding power output value for subsequent analysis and optimization. Record the start and end times and power range of each adjustable window in a standardized format to facilitate subsequent querying and analysis. Use visualization tools to graphically display the characteristics of adjustable power windows, making the power output of different windows clear at a glance. Use bar charts or line graphs to show power changes for each time period. The generated chart should be able to show the power changes for each adjustable window, helping to identify adjustment opportunities within different time periods.

[0015] Step S2: mining the trend of time-series energy flow and fitting the dynamic power distribution according to the multiple adjustable power windows to construct a dynamic power path distribution map; In this embodiment, the extracted adjustable power window data includes the time period, power output range, and related load conditions. For example, window 1 is 18 kW to 20 kW, with a time period of 30 minutes. The extracted data is cleaned to remove outliers and incomplete data points to ensure the accuracy of subsequent analysis. A threshold can be set to perform preliminary screening of power output to ensure data reliability. The normal fluctuation range of power output is set to ±2 kW, and abnormal records outside this range are eliminated. Appropriate trend analysis methods are selected, with common ones including time series analysis and moving average. These methods can identify long-term trends and short-term fluctuations in energy flow. The analysis time window is set to 15 minutes, and a moving average calculation is performed to smooth the data and reduce the impact of noise. Time series analysis is performed on the sorted adjustable power window data to calculate the energy flow changes within each window. This includes identifying rising, falling, and stable energy flow trends. For example, if the power within the window increases continuously over a period of time, it is recorded as an "upward energy flow trend," and vice versa. By analyzing the power data for each window, its changing characteristics are identified. Results Recording and Analysis: Trend analysis results are recorded in a database, including the trend type and duration for each window. This summary analysis can identify time periods that are most conducive to charging power adjustment. Fitting Model Selection: An appropriate model is selected to fit the dynamic power. Commonly used models include linear regression, nonlinear models, or polynomial fitting. The most appropriate fitting method is selected based on the data characteristics. For example, a polynomial fitting model is selected, with the fitting order set to 2 to capture the nonlinear characteristics of the data. Energy flow changes within each adjustable power window are fitted to generate a dynamic power distribution model. Fitting historical data can be used to predict future power distribution. For example, power change data for each window is fed into the fitting model, fitting parameters are calculated, and the corresponding dynamic equations are generated. Fitting results are recorded in a database, and visualization tools are used to compare the fitted curves with actual data. The resulting charts should intuitively demonstrate the model's fitting performance and the dynamic power distribution characteristics of each window. The charts should include an overlay of the actual power data points and the fitted curves to facilitate analysis of the fit performance. Path Distribution Analysis: Based on the fitting results, the dynamic paths of different adjustable power windows are analyzed. Identify the path characteristics of power output changes over time and construct a dynamic power path distribution map. For example, record the power change paths for each time period and identify which paths perform well under different load conditions. Integrate the dynamic power path distribution results obtained from the analysis into a map that displays the power output of each path in different time periods. Ensure that the map clearly conveys the power change trend and distribution characteristics. The map should include different colors or line types to represent the power change paths in different time periods to facilitate intuitive understanding of their dynamic characteristics.

[0016] Step S3: performing minimum power drift point detection according to the dynamic power path distribution map, and performing power fluctuation compensation inversion to obtain power fluctuation compensation configuration parameters; In this embodiment, the definition of the minimum power drift point is clarified. The drift point generally refers to the time point when the power output fluctuates slightly during the charging process. To this end, a detection standard is required, for example, the time point when the power fluctuation amplitude falls below a certain threshold (such as 2 kW) and lasts for more than 5 minutes.

[0017] A power fluctuation threshold of ±2 kW and a duration threshold of 5 minutes are set to identify stable drift points. Power time series data is extracted from the previously constructed dynamic power path distribution map in preparation for drift point detection. Data integrity and accuracy must be ensured for subsequent analysis. The extracted power data should include power values for each time period, recording power changes for at least one hour to facilitate comprehensive analysis. Drift point detection is implemented by using a sliding window algorithm to traverse the extracted power data and determine whether the power change at each time point meets the drift point detection criteria. If the power change amplitude continuously falls below the set threshold within a certain time period, the time point is recorded as a drift point. For example, if the power fluctuates from 20 kW to 18 kW within a certain period and remains within this range for more than 5 minutes, it is marked as a drift point. All detected minimum power drift points are recorded in a database, ensuring that each drift point has a corresponding timestamp and power value, forming a complete list of drift points. The criteria and objectives for power fluctuation compensation are determined. The goal is to analyze the characteristics of drift points and develop a power regulation strategy to ensure stable power output near the drift points. The compensation target is set to control the power output fluctuation near each drift point within ±1 kW to ensure the stability and efficiency of the charging process. Based on the power value and fluctuation characteristics of the drift point, the required power compensation parameters are calculated. This includes determining the power value that needs to be adjusted and the duration of the adjustment when the drift point occurs. For example, if the power at a drift point is 19 kW and the goal is to keep it between 18 kW and 20 kW, the compensation parameters should be set to 18 kW and 20 kW to ensure the stability of the charging power. Based on the calculated compensation parameters, a corresponding power adjustment strategy is formulated. This can be achieved by adjusting the control system settings to ensure that the charging power can be adjusted in time when the drift point is identified. Set the control system so that it can adjust the power to within the target range within 1 minute after the drift point is detected, and respond to changes in real time.

[0018] Step S4: performing local dynamic power regulation according to the power fluctuation compensation configuration parameters and constructing a local window dynamic power control strategy; In this embodiment, a local dynamic power adjustment strategy is designed based on power fluctuation compensation parameters. This strategy should flexibly adjust charging power based on real-time data feedback to ensure stability under varying load conditions. Set at a drift point, it monitors power changes in real time. If power output exceeds the target range, it immediately adjusts to maintain the set power range. A dynamic adjustment mechanism is established so that whenever a power drift point is detected, the control system can calculate the required power adjustment in real time and issue an adjustment command. The adjustment strategy is set to automatic, meaning that when the system detects power fluctuations near the drift point, it automatically makes compensation adjustments within a ±1 kW range. Real-time monitoring and feedback: During the charging process, the power output of the charging pile is monitored in real time and compared with the set target power range. If the power exceeds the range, the dynamic adjustment strategy is immediately implemented to make adjustments. Ensure that the system updates power output data every second and performs real-time comparisons to ensure rapid response. Based on the real-time monitoring results, local dynamic power adjustment is executed. If the power output falls below the target range, the charging power is increased; if it exceeds the target range, the charging power is reduced. This process should be completed within the set response time. For example, if the power is 21 kW at the drift point and the target is 20 kW, the system should adjust the power to 20 kW within 1 minute. Evaluate the effectiveness of the implemented local dynamic power regulation strategy, record the changes in power output before and after the adjustment, and analyze the effectiveness and stability of the regulation strategy. At each drift point, record the power values and fluctuations before and after the adjustment to evaluate the success rate of the adjustment. Based on the evaluation results, formulate corresponding optimization plans. If certain regulation strategies are found to be ineffective under specific load conditions, adjust them to improve the overall regulation effect. For example, if the regulation strategy fails to effectively control power fluctuations during certain high-load periods, consider increasing the regulation sensitivity or adjusting the target power range.

[0019] Step S5: performing charging simulation optimization according to the local window dynamic power control strategy, and performing power control feedback evaluation to generate power control feedback information; In this example, the total charging simulation duration is set to 2 hours, the initial charging power is 20 kW, and the monitoring window is set to 15 minutes. Ensure that sufficient data is collected within each window. During the charging simulation, key parameters, including charging power, current, voltage, and grid load status, are collected in real time. High-frequency data collection (e.g., every 5 seconds) is used to obtain more accurate information on power changes. Ensure that the data acquisition system can simultaneously record multiple parameters and quickly respond to power adjustment commands when needed. During the charging process, power output is monitored and adjusted based on the dynamic control strategy in the local window. If power is detected to be outside the target range at any time, the control command is immediately executed to adjust the power within the target range. For example, if the power output is 22 kW within a 15-minute window, exceeding the target value by 20 kW, the system should immediately adjust the power to 20 kW to maintain system stability. Feedback Information Collection: After the charging simulation is completed, power control feedback information is collected throughout the entire process. The response time and adjustment range of each power adjustment, as well as its impact on charging efficiency, are recorded. This ensures data integrity and accuracy. Record the timestamp of each adjustment, the power values before and after the adjustment, and the corresponding grid load for subsequent analysis. Determine indicators for power control feedback evaluation, including power fluctuation amplitude, charging efficiency, response time, etc. Use these indicators to comprehensively evaluate the effectiveness of the local dynamic power control strategy. Set the target charging efficiency to above 90%, control the power fluctuation amplitude within ±1 kW, and the response time should be less than 1 minute. Use statistical analysis methods to analyze the collected feedback information and evaluate whether each indicator meets the preset standards. By comparing the power output before and after the adjustment, identify the advantages and disadvantages of the control strategy. For example, the average power fluctuation amplitude throughout the charging process can be calculated and compared with the target value to determine the effectiveness of the control strategy.

[0020] Step S6: Perform global power control optimization and adaptive power feedback evolution according to the power control feedback information, and build an adaptive power optimization model.

[0021] In this embodiment, the power control feedback information obtained in the previous step is collected and organized. This information includes power fluctuation amplitude, charging efficiency, response time, and control results for each time period. Ensure that all data is cleaned and formatted for subsequent analysis. Experimental parameters: The feedback information is set to the power fluctuation records of the past 30 charging processes to ensure representative and comprehensive data. Select appropriate data analysis methods to evaluate the effectiveness of the feedback information. Statistical analysis, regression analysis, or other machine learning methods can be used to identify patterns and trends in the feedback information. Experimental parameters: A linear regression model is used to analyze the power control effectiveness, with the goal of identifying key factors affecting charging efficiency. Evaluate the feedback information and analyze the performance of the power control strategy under different load conditions. Identify situations where the control strategy performs well and where it needs improvement. For example, if the power fluctuation amplitude is generally large under high load conditions, record it as a "pending optimization" item and analyze the reasons for this. Based on the results of the feedback analysis, redesign the global power control strategy. The new strategy should comprehensively consider different scenarios and user needs to ensure efficient power allocation and control under various load conditions. Experimental Parameters: A new control strategy was established, including reducing the maximum power limit during high-load periods and allowing higher charging power during low-load periods. The new global power control strategy was implemented on a charging and discharging testbed. Multiple charging simulation experiments were conducted to verify the effectiveness and stability of the new strategy. This ensured that the collected data reflected the performance of the new strategy. Experimental Parameters: The simulation duration was set to 2 hours, and tests were conducted using different load variations (e.g., 10% and 20%) to assess the adaptability of the new strategy. The effectiveness of the implemented new strategy was evaluated, analyzing key indicators such as charging efficiency, power fluctuations, and user satisfaction. Based on the evaluation results, the control strategy was further optimized. For example, if charging efficiency during a certain period did not meet expectations, the power allocation strategy for that period was adjusted to ensure it better adapted to the load variation. Based on the results of the global control optimization, an adaptive power optimization model was constructed. This model should be able to learn and adjust the power control strategy in real time, automatically optimizing the charging strategy based on real-time data feedback. Experimental Parameters: The model learning cycle was set to update every 30 minutes to adapt the strategy to new data. During the charging process, power output and user feedback were monitored in real time and fed into the adaptive model. The model should automatically adjust its power regulation strategy based on the latest data to address load fluctuations. Experimental parameters: Ensure that the monitoring system collects power data once per second and provides timely feedback to the model when fluctuations are detected. Validate the adaptive power optimization model and evaluate its performance in real-world applications. Continuously optimize model parameters and regulation strategies based on user feedback and charging performance.

[0022] In this embodiment, refer to Figure 2, is a flowchart of the detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Define multiple grid load disturbance parameters; Using the test platform to simulate charging of a new energy charging pile, injecting dynamic load disturbances based on the multiple grid load disturbance parameters, and collecting full-cycle charging simulation data; Performing discrete calculation of charging power at multiple time points on the full-cycle charging simulation data to obtain a multi-time point charging power curve; Perform load disturbance analysis based on multi-time point charging power curves to identify power fluctuation characteristics under load disturbances; Locating a plurality of power fluctuation time points based on the power fluctuation characteristics; An adjustable power time period analysis is performed on the power fluctuation time point to generate a plurality of adjustable power windows.

[0023] In this embodiment, multiple grid load disturbance parameters are defined, including voltage step disturbance (within a ±10% range, simulating sudden voltage fluctuations); frequency disturbance (fluctuating between 49.5–50.5 Hz, simulating grid frequency offset); harmonic injection disturbance (3rd, 5th, and 7th harmonics, simulating nonlinear load interference); dynamic load impact disturbance (using a resistive-inductive composite load to achieve short-term impact fluctuations); and charge interruption / reconnection disturbance (simulating users frequently plugging and unplugging charging plugs). A test platform is used to establish a new energy charging pile charging simulation environment, ensuring that the equipment can simulate various grid load disturbances and charging conditions. The platform should have real-time data acquisition and monitoring capabilities. For example, it should be configured with charging piles, load simulators, and data acquisition systems to ensure that the platform can simulate various charging scenarios. During the charging simulation, dynamic load disturbance injection is performed based on the aforementioned load disturbance parameters. This can be achieved by programmatically controlling the load simulator to adjust the load state according to the set parameters. For example, a 10% load surge can be introduced at a certain charging stage, while the output power and other relevant parameters of the charging pile are continuously monitored. During the charging process, full-cycle charging simulation data, including charging power, voltage, current, and charging duration, is collected in real time to ensure data integrity and accuracy. For example, the power output of the charging station is recorded every minute to generate time series data for subsequent analysis. The collected full-cycle charging simulation data is processed and organized, arranged in chronological order, and formatted consistently for ease of subsequent calculations. For example, all data is imported into data analysis software, cleaned, and outliers and duplicate records are removed to ensure data reliability. Charging power is discretely calculated at multiple time points to generate a multi-point charging power curve. This can be achieved by extracting charging power data at different time points. For example, power calculations can be performed at 10-minute intervals, with the charging power at each time point recorded to form a power-time curve. Based on the multi-point charging power curve, load disturbance analysis is performed to identify power fluctuation characteristics. This can be achieved by calculating the power change rate or fluctuation amplitude. For example, the power change rate at each time point is calculated to identify fluctuation characteristics during load disturbances, and the fluctuation amplitude and frequency are recorded. The identified power fluctuation characteristics are recorded and categorized, ensuring that each characteristic has a corresponding description for subsequent analysis. For example, if a power fluctuation amplitude exceeding 5kW is recorded when the load suddenly changes, it will be marked as a "high fluctuation feature". Based on the power fluctuation characteristics, multiple power fluctuation time points are analyzed and located. This involves finding the time points in the power curve where the fluctuation amplitude exceeds the set threshold. For example, set the threshold to 5kW, scan the power curve, and record all time points that exceed the threshold to form a list of fluctuation time points. Perform feature analysis on the identified power fluctuation time points, and record the power value, fluctuation amplitude and related load disturbance conditions at each time point to gain a deeper understanding of the source of the fluctuation. For example, if the fluctuation amplitude at a certain time point is 7kW, record the load status when it occurs for subsequent analysis.The power fluctuation time points and their characteristics are recorded in the database and displayed through charts to help understand the characteristics of different fluctuation time points. For example, a time series graph is generated, all fluctuation time points are marked, and the occurrence of power fluctuations is visually displayed. According to the identified power fluctuation time points, the adjustable power time period is analyzed to determine when the charging power can be adjusted to cope with the fluctuation. For example, 10 minutes before and after the fluctuation time point is set as an adjustable time window to ensure that measures can be taken before and after the fluctuation occurs. Perform feature analysis on each adjustable power window and record the power changes and load status during these time periods in order to formulate corresponding adjustment strategies. For example, if the power change range is small within a certain adjustable window, you can consider adjusting the charging power within this time period.

[0024] In this embodiment, the specific steps of analyzing the adjustable power time period at the power fluctuation time point to generate multiple adjustable power windows are: Set the normal fluctuation amplitude and frequency to obtain the normal power change threshold; Performing abnormal power amplitude judgment on multiple power fluctuation time points according to the conventional power change threshold, and marking the power fluctuation time points exceeding the threshold; Calculate the threshold deviation power range at the power fluctuation time point; Defining an adaptive time span according to the threshold deviation power range to obtain an adaptive time span of a power fluctuation time point exceeding the threshold; The adjustable power time period is divided according to the adaptive time span to generate multiple adjustable power windows.

[0025] In this embodiment, a typical fluctuation amplitude and frequency are set for the charging pile. These parameters should be based on historical charging data and grid operating standards to ensure their rationality and reliability. For example, a typical fluctuation amplitude can be set to ±5 kW, with a frequency of once per minute. This means that under normal operating conditions, the charging power should fluctuate within this range. Statistical analysis of historical power data is performed to assess power fluctuations under normal operating conditions. Time series analysis can be used to identify the distribution characteristics of typical fluctuations. For example, by analyzing charging data from the past month, the average power fluctuation range under normal conditions can be calculated to confirm the rationality of the set parameters. Based on the set typical fluctuation amplitude and frequency, a typical power variation threshold is calculated. This threshold is recorded to facilitate subsequent judgment of power fluctuation time points. For example, a typical power variation threshold of ±5 kW can be set and recorded in the experimental documentation to provide a basis for subsequent analysis. Based on the set typical power variation threshold, a mechanism for determining abnormal power amplitude is established. This mechanism should be able to determine in real time whether each power fluctuation time point exceeds the set threshold. For example, if the charging power at a certain point in time is 30 kW, and the normal fluctuation range is 25 kW ± 5 kW, the power at that point in time is outside the normal range. Each power fluctuation time point is examined individually to determine whether it exceeds the threshold, and all abnormal time points are marked. This process can be automated using conditional statements or algorithms. For example, if the power at a certain time point is 32 kW, record that time point as "abnormal" and retain its power value and timestamp. Record the abnormality determination results in a database, ensuring that each abnormal time point has a corresponding status mark. Use visualization tools to display the distribution of abnormal time points. Based on the abnormality determination time points, calculate the threshold power deviation range for these time points. This involves determining the deviation of each abnormal time point from the normal fluctuation threshold. For example, if the power at an abnormal time point is 32 kW, and the normal fluctuation threshold is 25 kW ± 5 kW, the deviation is 32 kW - 30 kW = 2 kW. Calculate the deviation for each abnormal time point and record the absolute value and direction of the deviation. This helps better understand the severity of the power fluctuation and the nature of the deviation. For example, if the power deviation at a certain time point is a positive value of 2 kW, record it as a "positive deviation," and if it is a negative value, record it as a "negative deviation." Based on the threshold deviation power range, define an adaptive time span. This time span should reflect the impact of abnormal power fluctuations on the charging process and ensure that adjustments are made within critical time periods. For example, if the deviation power exceeds 2kW, the adaptive time span can be set to ±5 minutes to allow power adjustments during this period. For each abnormal time point, calculate the adaptive time span. The duration of similar fluctuations in historical data can be extrapolated to ensure the appropriateness of the time span.For example, if similar deviations are found in historical data that persist for three minutes, the adaptive time span for the current abnormal time point can be set to five minutes. Based on the calculated adaptive time span, adjustable power time periods are divided. Each adjustable time period should include the power fluctuations at the abnormal time point and within its adaptive time span. For example, if the adaptive time span for an abnormal time point is ±5 minutes, the power fluctuations at that time point and within the five minutes before and after that time point are considered an adjustable time period. The power data for each abnormal time point and its adaptive time span are grouped to generate multiple adjustable power windows. This can be automated through programming to ensure accurate division. For example, the start and end times of each adjustable time period, along with the corresponding power fluctuations, are recorded. The resulting adjustable power time periods are recorded in a database, and the distribution of all adjustable windows is displayed using a visual chart. A bar chart or timeline chart is generated to display each adjustable time period and its corresponding power fluctuations, facilitating subsequent optimization and adjustment strategies.

[0026] In this embodiment, refer to Figure 3 , is a flowchart of the detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Performing power interaction behavior detection on the multiple adjustable power windows and extracting real-time power interaction behavior monitoring data; Calculating the energy flow change rate within a period of the real-time power interaction behavior monitoring data; Mining the time series energy flow trend change of the energy flow change rate within the period to obtain the energy flow evolution trend; Identify charging power paths based on full-cycle charging simulation data and analyze multiple charging power paths in the power grid; performing elastic power path analysis on the plurality of charging power paths according to the plurality of adjustable power windows to identify the elastic power paths; The path distribution of the elastic power path is analyzed, and the dynamic power distribution is fitted to the energy flow evolution trend to construct a dynamic power path distribution map.

[0027] In this embodiment, a power interaction behavior detection mechanism is established to monitor the power output of charging piles and user charging behavior in real time. This mechanism should be able to record key data such as power changes, charging time, and user interaction frequency. For example, charging power data can be collected every second, recording the time each user starts and ends charging for subsequent analysis. Within an adjustable power window, the test platform monitors real-time power interaction behavior and collects power data. This includes charging power, number of user interactions, and power changes. For example, if the monitoring data within a certain adjustable window shows that the charging power increases from 20kW to 25kW, this change is recorded and saved as real-time monitoring data. Based on the real-time power interaction behavior monitoring data, a method for calculating the energy flow change rate is defined. The energy flow change rate can be expressed as the amount of power change per unit time. For example, the energy flow change rate calculation formula is set as: ΔE / Δt, where ΔE is the power change (kW) and Δt is the time period (hours). The collected real-time power data is periodically calculated to determine the energy flow change rate within each time period. This can be achieved by performing a sliding calculation within a set time window. For example, if the power increases from 20 kW to 25 kW over a period of 5 minutes, the energy flow rate of change is (25 - 20) / 5 = 1 kW / min. Determine methods for identifying energy flow trend changes, which may include time series analysis and smoothing techniques (such as moving averages) to identify long-term trends and short-term fluctuations. For example, you can use a moving average to smooth the energy flow rate of change to eliminate the impact of short-term fluctuations. Perform time series analysis on the calculated energy flow rate of change data to identify trends. This includes identifying rising, falling, and stable energy flow trends. For example, by analyzing the moving average of the energy flow rate of change, if a consistent upward trend is observed over a period of time, record it as an "upward energy flow trend." Use full-cycle charging simulation data to analyze multiple charging power paths in the power grid. Identify power output paths under different charging time periods and conditions. For example, analyze power output during different time periods during the charging process and record the start and end power values for each path. Analyze charging power paths using path identification algorithms (such as cluster analysis or graph theory) to identify the characteristics of multiple charging power paths. For example, if a path is found to have stable power output under high load conditions, record this path as a "high-load stable path." Analyze the elasticity characteristics of multiple charging power paths based on adjustable power windows. A flexible power path is one that maintains stable power output despite load fluctuations and demand changes. For example, the standard for a flexible path is to maintain power output within a range of ±5kW within a load change of ±10%. Perform elasticity analysis on the multiple identified charging power paths to evaluate their stability and adaptability under different load conditions.For example, if a path can still maintain power output within the set range under sudden load changes, it will be marked as an "elastic path". Determine the analysis method of path distribution, including statistical analysis and distribution model fitting, in order to identify the usage and distribution characteristics of different charging power paths. For example, set the normal distribution or other appropriate distribution model to fit the power output of the path. Dynamically fit the power distribution based on the collected charging power path data. Analyze the changes in power output in different time periods to identify the trend of power distribution. For example, use statistical software to fit the power output data, generate a power distribution model, and evaluate its goodness of fit. Record the results of path distribution and dynamic power distribution fitting in the database, and display the power path distribution map through visualization tools. For example, generate a heat map to show the usage frequency and output characteristics of different power paths to help understand the power distribution of charging piles.

[0028] In this embodiment, refer to Figure 4 , is a flowchart of the detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include: Calculating the power fluctuation amplitude of each path according to the dynamic power path distribution map to extract the power fluctuation amplitude of each path; Performing minimum power drift point detection according to the power fluctuation amplitude, and extracting multiple minimum power drift points; Performing energy consumption-efficiency dual balance calculations on each of the minimum power drift points, and marking the optimal point of power control benefit; The power fluctuation compensation inversion is performed on the optimal point of power regulation benefit to obtain the power fluctuation compensation configuration parameters.

[0029] In this embodiment, In this embodiment, step S4 includes the following steps: Calculate the transient load information of the charging pile and the power grid based on multiple adjustable power windows; Mining the time-varying characteristics of transient load information to generate the time-varying characteristics of transient load in each window; Perform similarity matching calculation on the optimal point of power regulation benefit according to the transient load time-varying characteristics, and mark an adjustable power window similar to the optimal point of power regulation benefit; The adjustable power window is locally dynamically power regulated according to the power fluctuation compensation configuration parameters, and a local window dynamic power regulation strategy is constructed.

[0030] In this embodiment, the calculation method for power fluctuation amplitude is clarified. Fluctuation amplitude is generally defined as the difference between the maximum and minimum power values along a path. This calculation can reflect the power stability of each path. For example, the calculation formula is set as: Fluctuation amplitude = Maximum power - Minimum power (in kW). Power output data for each path is extracted from the dynamic power path distribution map. The accuracy and completeness of the data are ensured to facilitate subsequent calculations. For example, for each path, its power output value during the charging cycle is recorded and a corresponding data sequence is constructed. The fluctuation amplitude of each extracted data is calculated for each path. This process can be automated using programming tools or statistical software to improve computational efficiency. For example, if the power output of a path is [20, 25, 22, 30, 27] kW, the fluctuation amplitude of the path is calculated to be 30 - 20 = 10 kW. The detection criteria for the minimum power drift point are defined. Drift points are generally characteristic points on the power output curve where small fluctuations occur. These points may indicate key locations for power regulation. For example, the drift point criteria can be set as the time point when the power fluctuation amplitude is less than a certain threshold (e.g., 2 kW) and lasts for more than 5 minutes. A detection algorithm can be designed to identify the minimum power drift point on each path. Sliding window analysis or threshold detection can be used to ensure detection accuracy. For example, the power data for each path is traversed. If the power fluctuation amplitude is less than 2 kW within a certain time period, this time point is recorded as a drift point. The identified minimum power drift points are recorded, ensuring that each drift point has a corresponding power value and timestamp for subsequent analysis. For example, the record format may include time point, power value, and path ID, forming a drift point database. Formulas for calculating energy consumption and efficiency are defined to facilitate evaluation of each minimum power drift point. Energy consumption generally refers to energy consumed within a certain time period, while efficiency can be expressed as the ratio of output power to input power. For example, energy consumption = power × time, and efficiency = output power / (output power + power loss). For each minimum power drift point, energy consumption and efficiency are calculated, and a dual balance analysis is performed. A loop can be used to iterate through all drift points and calculate the corresponding energy consumption and efficiency. For example, if the power at a drift point is 22 kW, the duration is 10 minutes, and the loss is 2 kW, the energy consumption is 22 kW × (10 / 60) = 3.67 kWh, and the efficiency can be calculated according to the formula. Based on the calculation results, mark the optimal points for power regulation benefits. These points are usually located at the locations with the lowest energy consumption and the highest efficiency. For example, if the energy consumption of a drift point is found to be 3.5 kWh and the efficiency is 90%, record this point as the "optimal point." The inversion process for defining power fluctuation compensation aims to determine how to adjust the power to compensate for fluctuations by analyzing the characteristics of the optimal point. For example, the goal of the compensation inversion is to keep the power output stable near the drift point to avoid unnecessary fluctuations.Based on the marked optimal point, the power fluctuation compensation configuration parameters are calculated. This involves determining the power value and duration required for adjustment within a specific time period. For example, if the power at the optimal point is 22 kW and the power needs to be maintained within ±1 kW around the drift point, the compensation parameters are 21 kW and 23 kW. Based on the calculated compensation parameters, a corresponding power adjustment strategy is developed. This can be achieved by programming the output power of the charging station to ensure timely adjustment when the drift point occurs. For example, the charging power can be adjusted to the compensation parameter range in real time within 5 minutes before and after the drift point. If the current power is 28 kW and the optimal point power is 25 kW, the system will generate an adjustment command of ΔPc = -3 kW. Configuration parameters include: compensation direction = reduction, amplitude = 3 kW, response delay = 1.2 s, and hold time = 15 s.

[0031] In this embodiment, step S5 includes the following steps: Perform charging simulation optimization based on the local window dynamic power control strategy and collect secondary charging simulation data; Defining a time length, decomposing the secondary charging simulation data into multiple time periods, and extracting charging simulation data of multiple time periods; Calculating the instantaneous power change amplitude of the charging simulation data to generate the instantaneous power change amplitude after power regulation; A power control feedback evaluation is performed on the instantaneous power variation amplitude based on the power fluctuation characteristics under the load disturbance to generate power control feedback information.

[0032] In this embodiment, a local window dynamic power control strategy is developed based on the previous power fluctuation analysis. This strategy aims to dynamically adjust the charging power according to the real-time grid load and charging demand to optimize charging efficiency and stability.

[0033] For example, charging power can be reduced during high-load periods (such as the evening rush hour) and increased during low-load periods (such as late at night) to balance grid load and improve charging station efficiency. The developed dynamic power control strategy is implemented on the charging test platform to conduct charging simulations. This involves dynamically adjusting the power output of the charging station according to the preset control strategy to ensure appropriate charging power during different time periods. For example, if the maximum charging power is limited to 15kW between 6:00 PM and 8:00 PM, the charging power is monitored and adjusted in real time during this period. During the charging simulation, secondary charging simulation data is collected in real time, including charging power, charging time, user interaction, and grid load conditions. This data provides a foundation for subsequent analysis. For example, minute-by-minute power output and grid status are recorded to ensure data integrity and accuracy for subsequent analysis. Time periods can be defined to decompose secondary charging simulation data into multiple time periods. The charging cycle can be divided into multiple time periods based on charging demand and grid load conditions. For example, by setting a time period of 15 minutes, the entire charging process can be divided into multiple 15-minute time periods to analyze charging conditions at different time periods. The collected secondary charging simulation data is broken down into predefined time periods, extracting the charging simulation data for each time period. This can be achieved through programming to automatically segment and organize the data. For example, if the total charging time is three hours, it can be divided into twelve 15-minute time periods, and the power data for each time period can be extracted. The charging simulation data for each time period is recorded in a database for subsequent analysis. Preliminary statistics are also performed to identify the charging characteristics of each time period. For example, a report can be generated summarizing the average charging power and total energy consumption for each time period to understand the charging efficiency of different time periods. A calculation method for the instantaneous power variation amplitude is defined. The instantaneous power variation amplitude generally indicates the maximum change in power output within a given time period. For example, the calculation formula can be set as: Instantaneous power variation amplitude = Maximum power - Minimum power. The instantaneous power variation amplitude is calculated for the charging simulation data for each time period. This can be achieved by iterating through the data for each time period and identifying the maximum and minimum power values. For example, if the power data for a certain time period is [10, 12, 15, 13, 11] kW, the instantaneous power variation is 15 - 10 = 5 kW. Based on the previous analysis of power fluctuation characteristics, the performance of the instantaneous power variation under load disturbances is evaluated. This requires considering the dynamic load conditions of the power grid and their impact on charging power. For example, the power output variation of the charging pile is evaluated under a grid load fluctuation of ±10%, and whether it can be maintained within a reasonable range. The instantaneous power variation is evaluated for power control feedback, identifying the power fluctuation characteristics under load disturbances and generating relevant feedback information.For example, if the instantaneous power change amplitude in a certain time period exceeds the set threshold (such as 5 kW), it will be recorded as "needs regulation" and marked with the reason and recommended adjustment strategy.

[0034] In this embodiment, step S6 includes the following steps: Based on the power control feedback information, the local window dynamic power control strategy is optimized for global power control and a global power control strategy is constructed; Perform cross-scenario transfer learning based on the global power control strategy to generate power control modes for multiple scenarios; Adaptive power feedback evolution is performed on the power control mode of the scenario to build an adaptive power optimization model.

[0035] In this embodiment, based on previous power control feedback, the performance of local window dynamic power control strategies in different scenarios is analyzed. The focus is on the amplitude and frequency of power fluctuations, as well as grid load conditions, to identify the strengths and weaknesses of local strategies. For example, the power stability and control effectiveness under high and low load conditions are recorded, and statistical metrics (such as standard deviation and mean) are used to evaluate the performance of each scenario. Based on the analysis results, a global power control strategy is designed. This strategy should comprehensively consider the needs of different scenarios to ensure efficient power allocation and regulation under various load conditions. For example, a dynamic power limit strategy can be implemented during high-load periods, while charging stations can be allowed to charge quickly during low-load periods. This strategy can effectively reduce the risk of grid load imbalance. The global power control strategy is implemented on a testbed, and its effectiveness and stability are verified through simulations. Various metrics are recorded and compared with local strategies to ensure the global strategy has better adaptability and efficiency. For example, the power output of the global strategy is tested under different load conditions to evaluate its performance under grid load fluctuations, and the strategy is continuously optimized based on the feedback. A cross-scenario transfer learning framework is established to facilitate the migration of the global power control strategy to different charging scenarios. The framework should include steps such as scenario feature extraction, model training, and policy transfer. For example, different charging scenarios (such as urban charging stations, residential communities, and shopping malls) should be selected, and characteristic indicators (such as user demand, charging power, and load type) should be defined for each scenario. Historical data from different charging scenarios should be collected and organized to construct a scenario dataset. The dataset should ensure diversity and representativeness to provide sufficient information for transfer learning. For example, data on charging habits, power requirements, and load variations should be collected for each scenario, and typical power control patterns for each scenario should be recorded. Based on the collected scenario data, a transfer learning model should be trained. By leveraging knowledge of global power control policies, the model should be optimized to adapt to new scenarios. For example, deep learning algorithms should be applied to the scenario data to ensure that the model understands the characteristics of power control in different scenarios and generates highly adaptable power control policies. Based on the transfer learning model, power control patterns for multiple scenarios should be generated. Each pattern should be tailored to the characteristics of a specific scenario to ensure good adaptability in real-world applications. For example, a peak-hour fast charging mode could be designed for urban charging stations, while a nighttime low-power mode could be designed for residential communities to ensure that charging needs in different scenarios are appropriately met. The generated scenario power control patterns should be validated on a testbed. By simulating actual charging scenarios, we evaluate the performance of each mode under different load conditions. For example, we simulate different charging scenarios during peak and off-peak periods, record key indicators such as power output, charging efficiency, and user satisfaction, and analyze the effectiveness of each mode. Based on the verification results, we optimize the scenario-specific power control mode. We adjust the parameter settings of each mode to improve its performance in different scenarios and ensure its ability to flexibly respond to unexpected situations.For example, if a certain mode performs poorly under high load, policy parameters can be adjusted, and different power limits or regulation methods can be tried to optimize performance. An adaptive power feedback mechanism should be established to enable timely adjustments to the power control mode based on changes in charging scenarios and user needs. This mechanism should be able to monitor power output and user feedback in real time. For example, power output data can be collected at regular intervals (e.g., 10 minutes) and analyzed in conjunction with user charging feedback (such as charging duration and power requirements). Using this collected data, an adaptive power optimization model is constructed. This model should be able to dynamically adjust the power control strategy based on historical data and real-time feedback, achieving continuous optimization. For example, machine learning algorithms can be used to analyze historical data to identify changing patterns in user charging behavior, allowing for real-time updates to the power control strategy. The adaptive power optimization model is validated on a testbed to evaluate its performance in real-world applications. Continuous iterative optimization based on model feedback ensures its long-term effectiveness. For example, changes in power output after applying the adaptive model in different scenarios can be recorded, and their impact on charging efficiency and grid load balance can be analyzed to ensure model effectiveness.

[0036] In this embodiment, a new energy charging simulation optimization device based on a charge and discharge test platform is provided, which is used to execute the new energy charging simulation optimization method based on the charge and discharge test platform as described above, including: A charging simulation module is used to simulate the charging of a new energy charging pile based on the charging and discharging test platform, analyze the adjustable power time period, and generate multiple adjustable power windows; A power path distribution module is used to mine the trend of time-series energy flow and fit the dynamic power distribution according to the multiple adjustable power windows, and to construct a dynamic power path distribution map; a fluctuation compensation module, configured to detect a minimum power drift point according to the dynamic power path distribution map and perform power fluctuation compensation inversion to obtain power fluctuation compensation configuration parameters; The local adjustment module is used to perform local dynamic power adjustment based on the power fluctuation compensation configuration parameters and build a local window dynamic power control strategy; The control feedback module is used to perform charging simulation optimization based on the local window dynamic power control strategy, and perform power control feedback evaluation to generate power control feedback information; The global control module is used to optimize global power control and adaptive power feedback evolution based on power control feedback information, and build an adaptive power optimization model The present invention analyzes the stability, volatility and other characteristics of the power curve during the charging process, identifies multiple adjustable power time periods, generates corresponding power adjustment windows, and provides boundary conditions and decision support for subsequent optimization operations. Avoid over-abstract modeling, achieve realistic simulation of the actual operating state, and improve the applicability and effectiveness of the power control strategy. Deeply explore the power change trajectory within each adjustable time window, capture detailed information such as the direction, amplitude, and frequency of the energy flow, and form time series trend data. Through mathematical modeling and fitting analysis, a "dynamic power path distribution map" is established to vividly display the changing process of power flow, providing a visual and quantifiable basis for refined control. Improve the understanding of the power fluctuation mechanism during the charging process and lay a data foundation for subsequent predictive control and adaptive adjustment. Accurately identify the critical point where power drift occurs during the charging process, which serves as the "grasp" of the adjustment strategy to achieve precise control. Introduce an inversion analysis mechanism to mathematically model the causes of fluctuations and back-derive parameters, and derive scientific and reasonable power fluctuation compensation configuration parameters. This system proactively suppresses potential large fluctuations, ensuring stable power output and improving charging equipment operational stability and power control sophistication. It flexibly implements power regulation within specific time windows and regions, avoiding resource waste and response delays associated with unified global control. It fine-grainedly divides control strategies based on real-time status and compensation parameters, improving control response speed and scenario adaptability. This helps reduce local load pressure, balance charging peak loads, and enhance the flexibility and practicality of the power control system. Backtesting and data feedback analysis are performed on local power control strategies, forming a closed-loop evaluation mechanism for control effectiveness. Dynamically collects information on the system's response to control strategies, such as the degree of fluctuation suppression, efficiency improvement, and anomaly response capabilities. This enables continuous iterative optimization of control strategies, providing decision data and strategy improvement directions for subsequent model upgrades. By integrating historical and real-time feedback data to form a multi-dimensional decision-making basis, global power strategy optimization is implemented, achieving intelligent control at the macro level. By incorporating machine learning and adaptive adjustment algorithms, the system is equipped with the capabilities of policy learning, autonomous evolution, and self-correction. The resulting adaptive power optimization model is portable and scalable, applicable to various new energy scenarios, enabling the transition from local control to global coordination. The system has the ability to continuously evolve and can effectively respond to challenges such as equipment diversification, rapid grid response, and complex charging behavior in new energy scenarios.

[0037] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0038] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A new energy charging simulation optimization method based on a charge and discharge test platform, characterized in that: The following steps are involved: Step S1: performing a new energy charging pile charging simulation based on the charging and discharging test platform, and performing an adjustable power time period analysis to generate multiple adjustable power windows; Step S2: mining the trend of time-series energy flow and fitting the dynamic power distribution according to the multiple adjustable power windows to construct a dynamic power path distribution map; Step S3: performing minimum power drift point detection according to the dynamic power path distribution map, and performing power fluctuation compensation inversion to obtain power fluctuation compensation configuration parameters; Step S4: performing local dynamic power regulation according to the power fluctuation compensation configuration parameters and constructing a local window dynamic power control strategy; Step S5: performing charging simulation optimization according to the local window dynamic power control strategy, and performing power control feedback evaluation to generate power control feedback information; Step S6: Perform global power control optimization and adaptive power feedback evolution according to the power control feedback information, and build an adaptive power optimization model.

2. The new energy charging simulation optimization method based on the charge and discharge test platform according to claim 1 is characterized in that: The specific steps of step S1 are: Define multiple grid load disturbance parameters; Using the test platform to simulate charging of a new energy charging pile, injecting dynamic load disturbances based on the multiple grid load disturbance parameters, and collecting full-cycle charging simulation data; Performing discrete calculation of charging power at multiple time points on the full-cycle charging simulation data to obtain a multi-time point charging power curve; Perform load disturbance analysis based on multi-time point charging power curves to identify power fluctuation characteristics under load disturbances; Locating a plurality of power fluctuation time points based on the power fluctuation characteristics; An adjustable power time period analysis is performed on the power fluctuation time point to generate a plurality of adjustable power windows.

3. The new energy charging simulation optimization method based on the charge and discharge test platform according to claim 2 is characterized in that: The specific steps of analyzing the adjustable power time period at the power fluctuation time point to generate multiple adjustable power windows are: Set the normal fluctuation amplitude and frequency to obtain the normal power change threshold; Performing abnormal power amplitude judgment on multiple power fluctuation time points according to the conventional power change threshold, and marking the power fluctuation time points exceeding the threshold; Calculate the threshold deviation power range at the power fluctuation time point; Defining an adaptive time span according to the threshold deviation power range to obtain an adaptive time span of a power fluctuation time point exceeding the threshold; The adjustable power time period is divided according to the adaptive time span to generate multiple adjustable power windows.

4. The new energy charging simulation optimization method based on the charge and discharge test platform according to claim 1 is characterized in that: The specific steps of step S2 are: Performing power interaction behavior detection on the multiple adjustable power windows and extracting real-time power interaction behavior monitoring data; Calculating the energy flow change rate within a period of the real-time power interaction behavior monitoring data; Mining the time series energy flow trend change of the energy flow change rate within the period to obtain the energy flow evolution trend; Identify charging power paths based on full-cycle charging simulation data and analyze multiple charging power paths in the power grid; performing elastic power path analysis on the plurality of charging power paths according to the plurality of adjustable power windows to identify the elastic power paths; The path distribution of the elastic power path is analyzed, and the dynamic power distribution is fitted to the energy flow evolution trend to construct a dynamic power path distribution map.

5. The new energy charging simulation optimization method based on the charge and discharge test platform according to claim 1 is characterized in that: The specific steps of step S3 are: Calculating the power fluctuation amplitude of each path according to the dynamic power path distribution map to extract the power fluctuation amplitude of each path; Performing minimum power drift point detection according to the power fluctuation amplitude, and extracting multiple minimum power drift points; Performing energy consumption-efficiency dual balance calculations on each of the minimum power drift points, and marking the optimal point of power control benefit; The power fluctuation compensation inversion is performed on the optimal point of power regulation benefit to obtain the power fluctuation compensation configuration parameters.

6. The new energy charging simulation optimization method based on the charge and discharge test platform according to claim 1 is characterized in that: The specific steps of step S4 are: Calculate the transient load information of the charging pile and the power grid based on multiple adjustable power windows; Mining the time-varying characteristics of transient load information to generate the time-varying characteristics of transient load in each window; Perform similarity matching calculation on the optimal point of power regulation benefit according to the transient load time-varying characteristics, and mark an adjustable power window similar to the optimal point of power regulation benefit; The adjustable power window is locally dynamically power regulated according to the power fluctuation compensation configuration parameters, and a local window dynamic power regulation strategy is constructed.

7. The new energy charging simulation optimization method based on the charge and discharge test platform according to claim 1 is characterized in that: The specific steps of step S5 are: Perform charging simulation optimization based on the local window dynamic power control strategy and collect secondary charging simulation data; Defining a time length, decomposing the secondary charging simulation data into multiple time periods, and extracting charging simulation data of multiple time periods; Calculating the instantaneous power change amplitude of the charging simulation data to generate the instantaneous power change amplitude after power regulation; A power control feedback evaluation is performed on the instantaneous power variation amplitude based on the power fluctuation characteristics under the load disturbance to generate power control feedback information.

8. The new energy charging simulation optimization method based on the charge and discharge test platform according to claim 1 is characterized in that: The specific steps of step S6 are: Based on the power control feedback information, the local window dynamic power control strategy is optimized for global power control and a global power control strategy is constructed; Perform cross-scenario transfer learning based on the global power control strategy to generate power control modes for multiple scenarios; Performing adaptive power feedback evolution on the power control mode of the scenario to construct an adaptive power optimization model; The global power control optimization is specifically as follows: calculating the real-time charging power of each charging gun in the current charging pile network and the charging current and voltage request value of the charging vehicle based on the power control feedback information; Performing demand sequence fitting based on the charging current and voltage request values to obtain a charging vehicle electrical characteristic demand sequence; Dynamically adjusting the real-time charging power based on a charging vehicle electrical characteristic demand sequence; The dynamic adjustment specifically includes: identifying the electrical characteristic requirement values of all vehicles; performing adaptive charging power gain processing on vehicles with high electrical characteristic requirement values; and performing charging power attenuation processing on vehicles with low electrical characteristic requirement values.

9. A new energy charging simulation optimization device based on a charging and discharging test platform, characterized in that: The method for implementing the new energy charging simulation optimization method based on the charging and discharging test platform according to claim 1 comprises: A charging simulation module is used to simulate the charging of a new energy charging pile based on the charging and discharging test platform, analyze the adjustable power time period, and generate multiple adjustable power windows; A power path distribution module is used to mine the trend of time-series energy flow and fit the dynamic power distribution according to the multiple adjustable power windows, and to construct a dynamic power path distribution map; a fluctuation compensation module, configured to detect a minimum power drift point according to the dynamic power path distribution map and perform power fluctuation compensation inversion to obtain power fluctuation compensation configuration parameters; The local adjustment module is used to perform local dynamic power adjustment based on the power fluctuation compensation configuration parameters and build a local window dynamic power control strategy; The control feedback module is used to perform charging simulation optimization based on the local window dynamic power control strategy, and perform power control feedback evaluation to generate power control feedback information; The global control module is used to perform global power control optimization and adaptive power feedback evolution based on power control feedback information, and to build an adaptive power optimization model.

Images