A dynamic voltage regulation charging control method for an electric two-wheeler charger

Abstract

The present application relates to the technical field of electric two-wheeled vehicle, especially to a dynamic voltage regulation charging control method for electric two-wheeled vehicle charger, which collects the voltage deviation of incoming line, cable interference, charging ripple and the actual and theoretical voltage of battery terminal in real time. By analyzing the voltage deviation fluctuation in historical period, the voltage regulation response hysteresis factor is calculated, and the error accumulation factor is obtained combined with the change of incoming line voltage; at the same time, according to the confusion degree and volatility of interference index and ripple, the interference voltage fluctuation mapping value is solved. Finally, the error accumulation factor, interference mapping value, hysteresis factor and their correlation are comprehensively considered to dynamically generate the voltage regulation compensation coefficient to correct the control strategy. This method can accurately quantify the voltage drop and surge influence caused by long cable, and adaptively overcome the system response lag and error accumulation, significantly improve the stability, response speed of electric two-wheeled vehicle charging voltage and the safety of electric two-wheeled vehicle battery under complex power grid environment.

Description

Technical Field

[0001] This invention relates to the field of electric two-wheeled vehicle technology, and in particular to a dynamic voltage regulation charging control method for electric two-wheeled vehicle chargers. Background Technology

[0002] With the acceleration of urbanization, the power infrastructure in many older communities is aging, with power lines generally suffering from problems such as long cables, small cross-sectional areas, and high contact resistance. In these communities, the demand for charging electric two-wheelers, widely used by residents, continues to grow. However, most charging devices rely on extension cords or power strips to draw power from indoors, resulting in the actual charging point being far from the transformer and a long power supply path. This long cable connection method easily causes severe voltage drops, especially during peak electricity consumption periods. Simultaneously, aging power grids have poor anti-interference capabilities and are susceptible to surge voltages and high-frequency interference caused by the start-up and shutdown of nearby high-power appliances, resulting in drastic input voltage fluctuations and high harmonic content, seriously affecting the stable operation of chargers.

[0003] Traditional electric two-wheeler chargers mostly employ a constant voltage and constant current charging mode with fixed parameters, and their control strategies lack the ability to sense and respond to disturbances in the upstream power grid. When the voltage drops, the charger may experience unstable output voltage due to insufficient input, prolonging charging time or even triggering protection shutdown. Conversely, under surge interference, instantaneous overvoltage may damage internal power devices or affect battery life. While some existing chargers incorporate simple voltage regulation circuits or PID feedback regulation, their control parameters are fixed and cannot be dynamically adjusted according to the complex and ever-changing power grid environment, making it difficult to balance response speed and system stability. Furthermore, due to the strong time-varying and nonlinear characteristics of cable impedance, load changes, and external interference, the actual output voltage during charging often lags behind the control command, creating a voltage regulation blind zone. This causes the battery terminal voltage to deviate from the ideal charging curve, affecting charging efficiency and battery health.

[0004] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0005] The main objective of this invention is to provide a dynamic voltage regulation charging control method for electric two-wheeled vehicle chargers. This method aims to solve the technical problems of existing electric two-wheeled vehicle chargers in old community long cable access scenarios, which lack adaptive sensing and dynamic compensation capabilities for voltage drop and surge interference. As a result, fixed parameter control is difficult to overcome system response lag and error accumulation, leading to unstable charging voltage, low efficiency and battery safety hazards.

[0006] To achieve the above objectives, the present invention provides a dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger, the method comprising: Real-time acquisition of the incoming voltage deviation, cable interference index, charging ripple amplitude, actual battery terminal charging voltage, and theoretical battery terminal charging voltage under voltage regulation control command of the electric two-wheeler charger at each moment; Based on the deviation fluctuation of the actual battery terminal charging voltage and the theoretical battery terminal charging voltage at each time point in the historical analysis period, the voltage regulation response hysteresis factor at each time point is obtained; based on the voltage regulation response hysteresis factor at the current time point and the degree of change of the incoming line voltage deviation at each time point in the historical analysis period, the error accumulation factor at the current time point is obtained. Based on the historical analysis of cable interference indicators, the degree of disorder and fluctuation of charging ripple amplitude, and the degree of change of incoming line voltage deviation at each time point within the historical analysis period, the interference voltage fluctuation mapping value at each time point is obtained. Based on the error accumulation factor, the interference voltage fluctuation mapping value, the voltage regulation response hysteresis factor at the current moment, and the correlation between the interference voltage fluctuation mapping value and the voltage regulation response hysteresis factor at each moment in the historical analysis period at the current moment, the voltage regulation compensation coefficient at the current moment is obtained; the voltage regulation compensation coefficient is used to perform dynamic voltage regulation charging control on the electric two-wheeler charger.

[0007] Optionally, obtaining the voltage regulation response hysteresis factor at each time step includes: The theoretical battery terminal charging voltage and the actual battery terminal charging voltage at all times within the historical analysis period of each time are arranged in time sequence to obtain the theoretical voltage sequence and the actual voltage sequence. The theoretical voltage sequence and the actual voltage sequence are time-series matched. The absolute value of the deviation between the actual battery terminal charging voltage in the actual voltage sequence and the theoretical battery terminal charging voltage matched in the theoretical voltage sequence is calculated. The absolute values ​​of the deviations corresponding to all actual battery terminal charging voltages in the actual voltage sequence are arranged to obtain the voltage error sequence at each time point. Calculate the variance of all elements within a preset sliding window for each element in the voltage error sequence, and denote it as the neighborhood deviation volatility of each element; arrange the neighborhood deviation volatility of all elements in the voltage error sequence to obtain the deviation volatility sequence. The deviation fluctuation sequence is integrated over the historical analysis period at each moment, and the ratio of the integration result to the duration of the historical analysis period is used as the voltage regulation response hysteresis factor at each moment.

[0008] Optionally, obtaining the error accumulation factor at the current moment includes: Based on the differences in the rate of change of incoming line voltage deviation between each time period and its neighboring time periods in the historical analysis of each time period, as well as the differences in incoming line voltage deviation between adjacent time periods, the voltage fluctuation chaos coefficient at each time period is obtained. Obtain the first-order difference sequence of the voltage error sequence at the current moment, and use the mean of the elements in the first-order difference sequence as the cumulative error intensity at the current moment. The error accumulation factor at the current moment is obtained based on the voltage fluctuation chaos coefficient, the voltage regulation response hysteresis factor, and the error accumulation intensity.

[0009] Optionally, obtaining the voltage fluctuation chaos coefficient at each moment includes: Obtain the maximum value point of the incoming line voltage deviation at all times within the historical analysis period for each time moment, and record the time corresponding to the maximum value point as the analysis time moment; Calculate the slope of the incoming line voltage deviation at each time point, and average the absolute difference between the slopes of the incoming line voltage deviation at each analysis time point and the slopes at other times in the neighboring time period to obtain the local voltage mutation degree at each analysis time point; the mean of the local voltage mutation degrees at all analysis times is denoted as the overall voltage mutation degree. The voltage fluctuation is obtained by averaging the absolute difference between the incoming line voltage deviations at each two adjacent analysis times. Based on the overall voltage abrupt change and the voltage fluctuation, the voltage fluctuation chaos coefficient at each moment is obtained.

[0010] Optionally, obtaining the interference voltage fluctuation mapping value at each moment includes: Cable interference index and charging ripple amplitude are recorded as analysis indicators; Curve fitting is performed on the same analytical index for all times within the historical analysis period for each time point to obtain two envelope lines of the obtained fitted curve; phase extraction is performed on the envelope lines to obtain phase values; the absolute difference of the phase values ​​of the two envelope lines of the fitted curve corresponding to each analytical index is recorded as the phase difference at each time point. Calculate the variance of each analytical indicator across all times within the historical analysis period for each time point, as the data volatility at each time point; Based on the phase difference and data volatility of all types of analytical indicators, the comprehensive interference degree at each moment is obtained; Based on the voltage fluctuation chaos coefficient and the comprehensive interference degree, the interference voltage fluctuation mapping value at each moment is obtained.

[0011] Optionally, obtaining the voltage regulation compensation coefficient at the current moment includes: The interference voltage fluctuation mapping value and the voltage regulation response hysteresis factor of all times in the historical analysis period of the current time are arranged in time sequence to obtain the fluctuation mapping sequence and the hysteresis factor sequence in turn. The correlation coefficient between the two sequences is obtained and recorded as the compensation confidence level at the current time. The product of the current disturbance voltage fluctuation mapping value and the voltage regulation response hysteresis factor is used as the disturbance severity. The voltage regulation compensation coefficient at the current moment is obtained based on the severity of the interference, the compensation confidence level, and the error accumulation factor.

[0012] Optionally, the dynamic voltage regulation charging control of the electric two-wheeler charger using the voltage regulation compensation coefficient includes: Obtain the ideal PID voltage regulation parameters of the charger, and use the sum of the constant 1 and the voltage regulation compensation coefficient at the current moment to weight the ideal PID voltage regulation parameters to obtain the corrected PID voltage regulation parameters at the current moment; use the corrected PID voltage regulation parameters to perform real-time dynamic voltage regulation charging control on the electric two-wheeler charger.

[0013] Optionally, obtaining the comprehensive interference degree at each time moment includes: Calculate the product of the phase difference and the data volatility for each analytical indicator at each time step, and use it as the local interference degree; sum the local interference degrees of all types of analytical indicators at each time step, and use it as the comprehensive interference degree at each time step.

[0014] Optionally, the least squares method can be used to perform curve fitting on the same analytical index for all times within the historical analysis period at each time point.

[0015] In addition, to achieve the above objectives, the present invention also provides a computer-readable storage medium storing a dynamic voltage regulation charging control program for an electric two-wheeled vehicle charger. When the dynamic voltage regulation charging control program for an electric two-wheeled vehicle charger is executed by a processor, it implements the steps of the dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger as described above.

[0016] This invention provides a dynamic voltage regulation charging control method for electric two-wheeler chargers. The method, by real-time sensing of multi-dimensional disturbance factors such as incoming line voltage deviation, cable interference, and charging ripple, constructs dynamic evaluation indicators such as voltage regulation response hysteresis factor, error accumulation factor, and interference voltage fluctuation mapping value. This accurately quantifies the impact of voltage drop and surge interference on the charging process under long cable access. Furthermore, based on multi-factor coupling analysis, a voltage regulation compensation coefficient is generated to dynamically correct PID control parameters. This effectively overcomes the problems of traditional chargers in complex power grid environments, such as slow response, large adjustment blind zone, and weak anti-interference capability. It significantly improves the stability and response accuracy of the charging voltage, ensuring the safety and efficiency of the charging process at the electric two-wheeler battery end, and extending the battery life of the electric two-wheeler. It is particularly suitable for application scenarios in old communities with harsh power grid conditions. Attached Figure Description

[0017] Figure 1This is a flowchart illustrating an embodiment of the dynamic voltage regulation charging control method for electric two-wheeled vehicle chargers of the present invention; Figure 2 This is a flowchart illustrating another embodiment of the dynamic voltage regulation charging control method for electric two-wheeled vehicle chargers of the present invention.

[0018] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0020] Reference Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of the dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger according to the present invention.

[0021] In one embodiment, the dynamic voltage regulation charging control method for the electric two-wheeled vehicle charger includes: Step S100: Real-time acquisition of the incoming voltage deviation, cable interference index, charging ripple amplitude, actual battery terminal charging voltage, and theoretical battery terminal charging voltage under voltage regulation control command of the electric two-wheeler charger at each moment.

[0022] The input voltage deviation can be the difference between the actual voltage at the charger input terminal and the rated grid voltage. It can be used to reflect the fluctuation of the power supply quality of the front-end grid and serve as a basic signal for disturbance sensing. In this embodiment, the input voltage deviation can be obtained by real-time acquisition of the input voltage through a voltage sampling circuit and comparison with a standard value. The cable interference index can be a quantitative parameter characterizing the comprehensive interference effects of high-frequency noise, impedance voltage drop, and nonlinear distortion introduced by long-distance power cables. It can be used to assess the influence of the cable path on the stability of the charging voltage. For example, the cable interference index can be derived based on the rate of change of current, voltage harmonic components, or impedance model. The charging ripple amplitude can be the peak value of the AC fluctuation component superimposed on the battery terminal charging voltage. It can be used to indicate the purity of the charging output and the filtering efficiency, reflecting the system's recovery capability after being disturbed. Furthermore, the charging ripple amplitude can be extracted by high-frequency sampling of the battery terminal voltage and band-pass or high-pass filtering. The actual battery terminal charging voltage can be the voltage value actually applied to the battery terminals under the combined effect of the current control command and grid disturbance. It can be used as a feedback signal to evaluate the control effect and deviation status. In one exemplary embodiment, the actual battery terminal charging voltage can be directly measured using an isolated voltage sensor to measure the voltage between the positive and negative terminals of the battery.

[0023] The theoretical battery terminal charging voltage can be the target battery terminal voltage that should be achieved under ideal, disturbance-free conditions, based on the current voltage regulation control command. It can be used as a reference benchmark to calculate control deviation and response hysteresis. In one specific embodiment, the theoretical battery terminal charging voltage can be generated by the controller based on a preset charging curve and the current stage command. The historical analysis period can be a continuous time window ending at the current moment, which can be used to provide the time series data basis required for dynamic factor calculation. For example, the historical analysis period can include, but is not limited to, one or more of short-time sliding windows, medium-cycle statistical windows, and long-cycle trend windows. Real-time acquisition of the electric two-wheeler charger's incoming line voltage deviation, cable interference index, charging ripple amplitude, actual battery terminal charging voltage, and theoretical battery terminal charging voltage under voltage regulation control command at each moment can be achieved by synchronously acquiring multiple voltage and interference-related signals in each control cycle and generating corresponding theoretical values. Furthermore, this operation can be achieved by using multi-channel ADC synchronous sampling combined with digital filters to extract each signal, or by using time-division multiplexing high-precision sampling circuits combined with timestamp alignment mechanisms to achieve quasi-synchronous acquisition, thereby providing a complete state-aware foundation for subsequent dynamic factor construction.

[0024] Step S200: Based on the deviation fluctuation of the actual battery terminal charging voltage and the theoretical battery terminal charging voltage at each moment within the historical analysis period, obtain the voltage regulation response hysteresis factor at each moment. Based on the voltage regulation response hysteresis factor at the current moment and the degree of change of the incoming line voltage deviation at each moment within the historical analysis period, obtain the error accumulation factor at the current moment.

[0025] The voltage regulation response hysteresis factor can be a dynamic indicator that quantifies the degree of delay and attenuation in the actual battery terminal voltage response after the voltage regulation control command is issued. It can be used to characterize the dynamic following capability of the control system in long cable scenarios. In this embodiment, the voltage regulation response hysteresis factor can be calculated based on the fluctuation amplitude and phase difference of the deviation between the actual and theoretical battery terminal voltages during the historical analysis period. Furthermore, the voltage regulation response hysteresis factor can be coordinated with the actual and theoretical battery terminal charging voltages, constructing its value through the fluctuation of their deviation; it also serves as an input variable for the error accumulation factor and the voltage regulation compensation coefficient. The error accumulation factor can be a dynamic indicator reflecting the degree of accumulation of systematic voltage deviation caused by the combined effect of historical incoming line voltage deviation and voltage regulation response hysteresis, and can be used to identify the trend risk of long-term deviation from the ideal charging curve. In a specific embodiment, the error accumulation factor can be weighted and integrated or averaged by combining the current voltage regulation response hysteresis factor with the changing trend of the incoming line voltage deviation during the historical analysis period. Furthermore, the error accumulation factor can rely on the historical data of the voltage regulation response hysteresis factor and the incoming line voltage deviation; and be output to the voltage regulation compensation coefficient calculation module.

[0026] Based on the deviation fluctuations between the actual and theoretical battery terminal charging voltages at each moment within the historical analysis period, the voltage regulation response hysteresis factor at each moment is obtained. This can be achieved by calculating the standard deviation, peak-to-peak value, or frequency domain energy distribution of the actual and theoretical voltage deviations within the historical period and mapping it to a hysteresis index. Further, this operation can be performed using the moving variance based on the deviation sequence, or by extracting the time-frequency localization features of the deviation signal through wavelet transform to characterize the hysteresis characteristics. This allows for the quantification of output lag caused by line impedance and device response limitations in the control system. Based on the voltage regulation response hysteresis factor at the current moment and the degree of change in the incoming line voltage deviation at each moment within the historical analysis period, the error accumulation factor at the current moment is obtained. This can be achieved by performing a weighted correlation operation between the voltage regulation response hysteresis factor and the historical incoming line voltage deviation sequence to generate a cumulative error estimate. In an exemplary embodiment, this operation can be performed by using an exponentially weighted moving average combined with hysteresis factor scaling, or by constructing a recurrent neural network model to predict the cumulative error trend. This allows for the identification of systemic undercharging or overcharging risks caused by persistent low voltage or frequent fluctuations.

[0027] Step S300: Based on the cable interference index, the degree of disorder and fluctuation of the charging ripple amplitude, and the degree of change of the incoming line voltage deviation at each time point in the historical analysis period, obtain the interference voltage fluctuation mapping value at each time point.

[0028] The interference voltage fluctuation mapping value can be a dynamic quantitative value that integrates cable interference indicators, charging ripple amplitude, and incoming voltage deviation fluctuation characteristics to map the comprehensive impact of external disturbances on the charging process. This value can be used to uniformly characterize the coupled impact strength of multi-source interference on voltage stability. In this embodiment, the interference voltage fluctuation mapping value can be obtained by nonlinearly mapping or weighted synthesis of the disorder (e.g., entropy) and fluctuation amplitude (e.g., standard deviation) of three types of disturbance signals during historical analysis periods. Furthermore, the interference voltage fluctuation mapping value can be jointly driven by cable interference indicators, charging ripple amplitude, and incoming voltage deviation; participate in the generation of voltage regulation compensation coefficients; and undergo correlation analysis with the voltage regulation response hysteresis factor.

[0029] Based on the historical analysis of cable interference indicators, the degree of disorder and fluctuation of charging ripple amplitude, and the degree of change in incoming line voltage deviation at each time point within a given historical period, an interference voltage fluctuation mapping value is obtained for each time point. This can be achieved by calculating the information entropy (disorder) and standard deviation (fluctuation) of the three types of disturbance signals separately, and then fusing them into a single mapping value using a nonlinear function. Furthermore, this operation can be further refined by using a fuzzy logic system to perform rule-based fusion of multidimensional disturbance features, or by using principal component analysis to reduce dimensionality and then weighted to synthesize the mapping value, thereby enabling a unified representation of the comprehensive destructive force of composite interference on charging stability.

[0030] Step S400: Based on the error accumulation factor, interference voltage fluctuation mapping value, voltage regulation response hysteresis factor at the current moment, and the correlation between the interference voltage fluctuation mapping value and the voltage regulation response hysteresis factor at each moment in the historical analysis period, obtain the voltage regulation compensation coefficient for the current moment. Use the voltage regulation compensation coefficient to perform dynamic voltage regulation charging control on the electric two-wheeler charger.

[0031] The voltage regulation compensation coefficient can be a proportional adjustment factor generated based on multi-dimensional dynamic evaluation indicators and their historical correlations. It is used to correct control parameters and dynamically adjust PID control parameters to adapt to the current power grid disturbance state, improving voltage regulation accuracy and response speed. In this embodiment, the voltage regulation compensation coefficient can be calculated by coupling the current error accumulation factor, the interference voltage fluctuation mapping value, and the voltage regulation response hysteresis factor, and combining the time correlation of the latter two (such as cross-correlation coefficients or covariance) over historical periods. Furthermore, the voltage regulation compensation coefficient can receive three major dynamic factors as inputs; the output is used to dynamically correct PID control parameters, thereby affecting the voltage regulation behavior of the electric two-wheeled vehicle charger.

[0032] An electric two-wheeler charger can be a power electronic device that provides energy conversion and charging management for the electric two-wheeler's power battery. It can be used to execute a dynamic voltage regulation charging control strategy to achieve safe and efficient charging. In one specific embodiment, the electric two-wheeler charger may include an AC-DC converter, a DC-DC regulator, a control unit, and a sensing feedback loop. The dynamic voltage regulation charging control can be a closed-loop control strategy that dynamically adjusts the output voltage based on real-time grid disturbances and battery status. This can be used to ensure that the battery terminal voltage closely tracks the ideal charging curve, overcoming the adverse effects of long cables and interference. The PID control parameters can be adjustable coefficients in a proportional-integral-derivative controller, which can directly affect the speed of charging voltage regulation, overshoot, and steady-state accuracy. In an exemplary embodiment, the PID control parameters are fixed values ​​in conventional schemes; in this scheme, they are dynamically corrected by a voltage regulation compensation coefficient.

[0033] Based on the current error accumulation factor, interference voltage fluctuation mapping value, voltage regulation response hysteresis factor, and the correlation between the interference voltage fluctuation mapping value and the voltage regulation response hysteresis factor at each time point within the historical analysis period, the voltage regulation compensation coefficient for the current moment can be obtained. This can be achieved by using the three dynamic factors as the main inputs and introducing the cross-correlation coefficient or covariance of their historical sequences as coupling weights, and generating the compensation coefficient through a preset mapping function. Furthermore, this operation can achieve multivariate mapping by using a lookup table combined with interpolation, or by deploying a lightweight regression model (such as a linear combination + saturation function) to calculate the compensation coefficient online, thereby generating an adaptive adjustment factor that combines current disturbance intensity perception with understanding of system dynamic characteristics.

[0034] Dynamic voltage regulation charging control of electric two-wheeler chargers using a voltage regulation compensation coefficient can be achieved by applying the compensation coefficient to the charger's voltage regulation circuit to adjust the output setpoint or control gain. In one specific embodiment, this operation can generate a corrected target voltage by multiplying the compensation coefficient by the theoretical voltage, or by using the compensation coefficient to scale the output limiting or feedforward gain of the PID controller. This enables active compensation for grid disturbances and improves the battery terminal voltage tracking accuracy. Dynamically correcting the PID control parameters can be achieved by adjusting the proportional, integral, or derivative coefficients of the PID controller in real time based on the voltage regulation compensation coefficient. Furthermore, this operation can be performed by adjusting only the proportional gain Kp for a rapid response to disturbances, or by simultaneously adjusting Ki and Kd to balance steady-state error and overshoot suppression. This enhances system robustness under strong disturbances, improves regulation accuracy during stable periods, and avoids performance compromises caused by fixed parameters.

[0035] Taking the peak nighttime charging period in older communities as an example, the dynamic voltage regulation charging control method for electric two-wheeled vehicles in this embodiment addresses the situation where multiple households simultaneously use extension cords to charge their electric two-wheeled vehicles during peak electricity consumption, causing the incoming line voltage to drop below 190V. Simultaneously, the starting and stopping of nearby air conditioners triggers surges. The charger collects significant incoming line voltage deviations, high charging ripple, and severe cable interference in real time. Based on this, the system calculates a high voltage regulation response hysteresis factor (due to slow output response caused by long cables), a high error accumulation factor (due to insufficient charging caused by continuous low voltage), and a high interference voltage fluctuation mapping value (due to the superposition of surges and harmonics). These three factors are coupled to generate a large voltage regulation compensation coefficient, dynamically increasing the PID proportional gain and raising the target voltage, ensuring that the actual battery terminal voltage remains at the level required for the constant current stage, avoiding protection shutdown, and ensuring continuous completion of the charging process.

[0036] In one embodiment, obtaining the voltage regulation response hysteresis factor at each time step includes: The theoretical battery terminal charging voltage and the actual battery terminal charging voltage at all times within the historical analysis period at each time point are arranged in time sequence to obtain the theoretical voltage sequence and the actual voltage sequence.

[0037] The theoretical voltage sequence can be an ordered data column formed by arranging the theoretical battery terminal charging voltage at each moment within the historical analysis period in chronological order. It can be used as a time trajectory benchmark for the ideal control output, for comparison and analysis with the actual response. In this embodiment, the theoretical voltage sequence can be extracted and sorted by timestamp from the target voltage values ​​corresponding to the voltage regulation command recorded internally by the controller at each moment. Exemplarily, the theoretical voltage sequence can be used in conjunction with the actual voltage sequence to form the basic input for dynamic error analysis. The actual voltage sequence can be an ordered data column formed by arranging the actual battery terminal charging voltage at each moment within the historical analysis period in chronological order. It can be used to reflect the true dynamic response process of the system under disturbance conditions. Further, the actual voltage sequence can be formed by aligning and sorting historical data collected by voltage sensors according to timestamps. In an exemplary embodiment, the actual voltage sequence and the theoretical voltage sequence jointly participate in the subsequent timing alignment and error quantization process.

[0038] The theoretical and actual battery terminal charging voltages for all moments within the historical analysis period at each moment are arranged chronologically to obtain the theoretical voltage sequence and the actual voltage sequence. This can be achieved by extracting the two types of voltage data within a specified time window from a storage buffer and sorting them in ascending order based on timestamps. Furthermore, this operation can be implemented by using a circular buffer to maintain the latest N sampling points and directly outputting an ordered sequence, or by reading time-stamped data from non-volatile storage and then sorting it, thereby constructing a structured input sequence that can be used for dynamic comparative analysis.

[0039] The theoretical voltage sequence and the actual voltage sequence are time-series matched. The absolute value of the deviation between the actual battery terminal charging voltage in the actual voltage sequence and the theoretical battery terminal charging voltage matched in the theoretical voltage sequence is calculated. The absolute values ​​of the deviations corresponding to all actual battery terminal charging voltages in the actual voltage sequence are arranged to obtain the voltage error sequence at each time point.

[0040] Timing matching can be an operation that aligns corresponding time points or causal correlation points in a theoretical voltage sequence with those in an actual voltage sequence. It can be used to ensure that deviation calculations are based on the correct correspondence between the same control command and its implementation result. In this embodiment, timing matching can achieve a one-to-one mapping between elements of two sequences through timestamp alignment, delay compensation estimation, or dynamic time warping algorithms. For example, timing matching can work in conjunction with deviation absolute value calculation to ensure the causal consistency of error indicators. The deviation absolute value can be the absolute value of the difference between the actual battery terminal charging voltage and the corresponding theoretical battery terminal charging voltage after timing matching. It can be used to eliminate the influence of error direction and focus on quantifying the degree of response deviation. Furthermore, the deviation absolute value can be obtained by subtracting each pair of matched voltage values ​​and taking the absolute value. In a specific embodiment, the deviation absolute value serves as an unsigned error intensity indicator, supporting the construction of subsequent error sequences. Calculating the absolute value of the deviation between each actual battery terminal charging voltage in the actual voltage sequence and the matched theoretical battery terminal charging voltage in the theoretical voltage sequence can be achieved by subtracting each pair of matched voltage values ​​and taking the absolute value. Furthermore, this operation can be accelerated by using a lookup table in a fixed-point processor to calculate the absolute value, or by directly calling the fabs function in a floating-point unit, thereby generating an unsigned error strength index that focuses on the magnitude of the deviation rather than its direction.

[0041] A voltage error sequence can be a sequence composed of the absolute values ​​of all deviations within a historical analysis period, arranged in chronological order. It can be used to characterize the output tracking error intensity of the control system at each moment within the historical period. In this embodiment, the voltage error sequence can be arranged into a one-dimensional array by arranging all the absolute values ​​of deviations according to their occurrence time. For example, the voltage error sequence serves as the raw input for local fluctuation analysis, used in conjunction with a preset sliding window. Arranging the absolute values ​​of deviations corresponding to all actual battery terminal charging voltages in the actual voltage sequence yields the voltage error sequence for each moment. This can be achieved by organizing the calculated absolute values ​​of deviations into a one-dimensional array according to their corresponding time order. Furthermore, this operation can be implemented by directly filling the error array using the original time index or by creating a new independent buffer to store the error sequence, thereby forming a continuous error evolution trajectory and supporting subsequent local and global statistical analysis.

[0042] Calculate the variance of all elements within a preset sliding window for each element in the voltage error sequence, and denote it as the neighborhood deviation volatility of each element; arrange the neighborhood deviation volatility of all elements in the voltage error sequence to obtain the deviation volatility sequence.

[0043] The preset sliding window can be a fixed-length continuous sub-interval defined on the voltage error sequence for local statistical analysis. It can be used to limit the neighborhood range to calculate local fluctuation characteristics and suppress isolated noise interference.

[0044] Neighborhood deviation volatility can be the variance of all elements within a preset sliding window of a voltage error sequence. It can be used to measure the dispersion of the error within a local time period, reflecting the instantaneous change in response stability. Furthermore, neighborhood deviation volatility can be used to calculate the sample variance of the subsequence covered by the sliding window, serving as a volatility index for the center or end point. In a specific embodiment, neighborhood deviation volatility works in conjunction with the voltage error sequence to achieve a feature transformation from the original error to dynamic instability. Calculating the variance of all elements within a preset sliding window of each element in the voltage error sequence, denoted as the neighborhood deviation volatility of each element, can be achieved by sliding a fixed-length window across the voltage error sequence and calculating the variance of the data within the window at each position. Further, this operation can reduce computational complexity by employing a recursive variance algorithm or by using an FPGA to compute the variance of multiple windows in parallel, thereby extracting the local instability features of the error signal and distinguishing between sustained offsets and random jitter.

[0045] The deviation fluctuation sequence can be a sequence formed by arranging the neighborhood deviation fluctuation degrees corresponding to each element in the voltage error sequence according to their original time order. It can be used to transform the original error into a higher-order feature sequence describing its dynamic instability. In this embodiment, the deviation fluctuation sequence can traverse the voltage error sequence, calculate the neighborhood deviation fluctuation degree point by point, and then reassemble it into a new sequence. For example, the deviation fluctuation sequence serves as the input to the integration operation, carrying the evolution information of the response fluctuations over historical periods. Arranging the neighborhood deviation fluctuation degrees of all elements in the voltage error sequence to obtain the deviation fluctuation sequence can be achieved by arranging the neighborhood deviation fluctuation degrees calculated at each position into a new sequence according to their original time order. Furthermore, this operation can be implemented by replacing the voltage error sequence in place to save memory, or by allocating a new buffer to store the deviation fluctuation sequence, thereby transforming the original error into a higher-order representation describing its dynamic fluctuation characteristics.

[0046] The deviation fluctuation sequence is integrated over the historical analysis period at each moment, and the ratio of the integration result to the duration of the historical analysis period is used as the voltage regulation response hysteresis factor at each moment.

[0047] The integration result can be the cumulative amount obtained by numerically integrating the deviation fluctuation sequence over the historical analysis period, and can be used to summarize the total response instability over the entire historical period. In this embodiment, the integration result can be discretely integrated on the deviation fluctuation sequence using the trapezoidal method, rectangular method, or Simpson's method. For example, the integration result and the duration of the historical analysis period are used together for normalization to generate a standardized hysteresis index. Integrating the deviation fluctuation sequence over the historical analysis period at each moment can be done by performing numerical integration on the interval covered by the historical analysis period. Furthermore, this operation can be achieved by using an accumulator to sum the discrete points for approximate integration, or by using a weighted numerical integration method to improve accuracy, thereby accumulating the total response instability over the entire analysis window and reflecting the overall hysteresis trend.

[0048] The duration of the historical analysis period can be the time interval between the current moment and the starting point of its historical analysis. This interval can be used as a normalization factor to give the voltage regulation response hysteresis factor the meaning of average fluctuation per unit time. In this embodiment, the duration of the historical analysis period can be obtained by subtracting the start and end times recorded by the system clock. The ratio of the integral result to the duration of the historical analysis period is used as the voltage regulation response hysteresis factor at each moment. This can be achieved by dividing the integral result by the duration of the historical analysis period to obtain the average fluctuation per unit time. The formula for calculating the voltage regulation response hysteresis factor can be as follows:

[0049] In the formula, The voltage regulation response hysteresis factor is used to comprehensively evaluate the response hysteresis of the charger throughout the entire historical period. The larger the value, the slower the system response and the worse the stability. The neighborhood deviation fluctuation is given by the sliding window variance of the voltage error sequence. The calculation yields a result that characterizes the local jitter of the error. The historical analysis period serves as a time benchmark, used to average volatility over time to obtain a factor reflecting the overall level. Furthermore, this operation can also be achieved through approximate division via shifting in fixed-point arithmetic, or by directly performing division in a floating-point system, thereby generating a standardized hysteresis index with physical comparability, facilitating comparisons across time periods or devices.

[0050] For example, in a scenario where the evening charging peak in an older community is accompanied by interference from air conditioner start-stop, the dynamic voltage regulation charging control method for electric two-wheeled vehicles in this embodiment can be as follows: A user starts charging at 18:30. At this time, the grid voltage has dropped due to multiple households using electricity, and the frequent start-stop of neighboring air conditioners causes surges. The charger calculates the voltage regulation response hysteresis factor at t=18:35: First, the theoretical and actual battery terminal voltages of the past 5 minutes are extracted and a voltage error sequence is generated after time-series matching; it is found that the error increases sharply at the moment the air conditioner starts but quickly falls back. The neighborhood deviation fluctuation is calculated through a preset sliding window (e.g., 10 seconds), and it is identified that the instantaneous disturbance does not cause continuous instability; after integrating the deviation fluctuation sequence and dividing it by 300 seconds, a low voltage regulation response hysteresis factor is obtained, indicating that although the system is disturbed, its response capability has not significantly degraded. Therefore, the voltage regulation compensation coefficient is only slightly adjusted to avoid over-correction and oscillation.

[0051] In one embodiment, obtaining the error accumulation factor at the current moment includes: Based on the differences in the rate of change of incoming line voltage deviation between each time period and its neighboring time periods within the historical analysis period, as well as the differences in incoming line voltage deviation between adjacent time periods, the voltage fluctuation chaos coefficient for each time period is obtained.

[0052] The neighborhood time period can be a local time window surrounding a specific moment within the historical analysis period, used to calculate local dynamic characteristics. In this embodiment, the neighborhood time period can be used to provide a local reference benchmark for the rate of change of the incoming line voltage deviation, supporting chaotic analysis. The difference in the rate of change of the incoming line voltage deviation can be the difference between the rate of change of the incoming line voltage deviation at a certain moment within the historical analysis period and the average rate of change within its neighborhood time period. In an exemplary embodiment, the difference in the rate of change of the incoming line voltage deviation can be used to reflect the degree of abrupt change in the local voltage change trend, used to identify non-stationary disturbances. For example, the difference in the rate of change of the incoming line voltage deviation can be obtained by calculating the derivative or difference of the incoming line voltage deviation between adjacent sampling points, and then comparing it with the neighborhood mean. The difference in the incoming line voltage deviation between adjacent moments can be the difference in the incoming line voltage deviation between two consecutive sampling moments within the historical analysis period. In a specific embodiment, the difference in the incoming line voltage deviation between adjacent moments can be used to characterize the instantaneous jump intensity of voltage disturbances, capturing surge or drop events. Further, the difference in the incoming line voltage deviation between adjacent moments can be obtained by performing a first-order difference operation on the incoming line voltage deviation sequence.

[0053] The voltage fluctuation chaos coefficient is a quantitative indicator that characterizes the nonlinearity and unpredictability of power grid disturbances, constructed based on the difference in the rate of change of incoming line voltage deviation and the difference in deviation between adjacent moments. In this embodiment, the voltage fluctuation chaos coefficient can be used to identify sudden, non-periodic voltage disturbances caused by the start-up and shutdown of high-power appliances in old communities. For example, the voltage fluctuation chaos coefficient can be generated by fusing two types of difference indicators through nonlinear mapping (such as entropy value, Lyapunov exponent approximation, or variance ratio). Based on the difference in the rate of change of incoming line voltage deviation between each moment in the historical analysis period and its neighboring periods, as well as the difference in incoming line voltage deviation between adjacent moments, the voltage fluctuation chaos coefficient for each moment can be obtained. This can be achieved by calculating the difference between the rate of change of incoming line voltage deviation and the mean of the neighborhood for each moment in the historical analysis period, as well as the difference in deviation from the previous moment, and then fusing them into a chaos coefficient through a preset function. Furthermore, this operation can be achieved by quantifying the distribution uncertainty of the two types of differential sequences using information entropy, or by using the normalized variance ratio to measure the intensity of local mutations and weighted synthesis. This can effectively identify nonlinear and sudden components in power grid disturbances and improve the ability to perceive the complexity of external disturbances.

[0054] Obtain the first-order difference sequence of the voltage error sequence at the current moment, and use the mean of the elements in the first-order difference sequence as the cumulative error intensity at the current moment.

[0055] A first-order difference sequence can be a new sequence obtained by subtracting adjacent elements from the voltage error sequence at the current moment. In an exemplary embodiment, the first-order difference sequence can be used to reveal the direction and rate of voltage error evolution over time, distinguishing between static offset and dynamic accumulation. Obtaining the first-order difference sequence of the voltage error sequence at the current moment can be achieved by performing adjacent element subtraction on the voltage error sequence corresponding to the current moment to generate the difference sequence. Furthermore, this operation can be implemented by using a circular buffer to calculate the difference in-place in an embedded system, or by accelerating vector difference operations through DSP instructions, thereby converting static error into a dynamic trend signal, facilitating the identification of persistent offset.

[0056] The cumulative error intensity can be the mean of all elements in the first-order difference sequence, reflecting the overall growth or decay trend of the voltage error. In a specific embodiment, the cumulative error intensity can be used to quantify whether there is a continuous accumulation of error in the same direction in the system, avoiding misjudgment caused by relying solely on instantaneous errors. Furthermore, the cumulative error intensity can be obtained by calculating the arithmetic mean of the first-order difference sequence. Using the mean of the elements in the first-order difference sequence as the cumulative error intensity at the current moment can be achieved by averaging all elements in the first-order difference sequence, serving as a scalar indicator of the error evolution trend. Furthermore, this operation can be achieved by using the difference mean after moving average filtering to suppress noise, or by introducing a confidence interval to determine whether the mean deviates significantly from zero. Thus, if the mean is significantly greater than zero, it indicates that the error is continuously increasing, and there is a risk of systematic under-adjustment or over-adjustment.

[0057] The error accumulation factor at the current moment is obtained based on the voltage fluctuation chaos coefficient, voltage regulation response hysteresis factor, and error accumulation intensity.

[0058] This operation can take three indicators as input variables and generate the final factor through weighted fusion, table lookup interpolation, or lightweight model mapping. Furthermore, this operation can be achieved by using a fuzzy inference system to perform rule-based fusion of the three types of indicators, or by using a linear combination with a saturation function to achieve smooth mapping, thereby generating a high-dimensional cumulative error representation that comprehensively considers the chaos of external disturbances, the lag in system response, and the error development trend.

[0059] For example, in a scenario where air conditioners in older communities experience concentrated start-stop interference at night, the dynamic voltage regulation charging control method for electric two-wheeled vehicles in this embodiment can address this issue. During the peak electricity consumption period at 20:00, multiple air conditioner compressors start and stop simultaneously, causing frequent voltage fluctuations in the incoming line. The charger calculates the error accumulation factor at t=20:05: First, based on historical 5-minute data, it is found that at a certain moment, the incoming line voltage deviation jumps from -10V to +8V within 1 second, with extremely large differences between adjacent voltages, and its rate of change significantly deviates from the neighborhood mean, generating a high voltage fluctuation chaos coefficient. Simultaneously, the voltage error sequence shows that the actual battery terminal voltage is consistently lower than the theoretical value, and its first-order difference mean is positive, indicating continuous error accumulation. Combined with the previously calculated high voltage regulation response hysteresis factor (due to slow response caused by long cables), these three factors are integrated to generate a high error accumulation factor, triggering the voltage regulation compensation coefficient to increase the target voltage and enhance the integral effect, compensating for potential undercharging risks in advance and preventing the battery from being in a low-voltage charging state for extended periods.

[0060] In one embodiment, obtaining the voltage fluctuation chaos coefficients at each time step includes: Obtain the maximum value point of the incoming line voltage deviation at all times within the historical analysis period for each time moment, and record the time corresponding to the maximum value point as the analysis time moment; Calculate the slope of the incoming line voltage deviation at each time point, and average the absolute difference between the slopes of the incoming line voltage deviation at each analysis time point and the slopes at other times in the neighboring time period to obtain the local voltage change rate at each analysis time point; the mean of the local voltage change rates at all analysis times is denoted as the overall voltage change rate. The voltage fluctuation is obtained by averaging the absolute difference between the incoming line voltage deviations at each two adjacent analysis times. Based on the overall voltage abrupt change and voltage fluctuation, the voltage fluctuation chaos coefficient at each moment is obtained.

[0061] The maximum point can be the sampling point corresponding to a local maximum in the incoming voltage deviation sequence within the historical analysis period. It can be used to identify typical moments of most severe voltage disturbances, serving as key anchor points for analyzing abrupt events. In an exemplary embodiment, the maximum point can be identified as a local peak using sliding window comparison or first derivative sign detection. Furthermore, a one-to-one mapping relationship can be established between the maximum point and the analysis time. For example, the identification of the maximum point can be combined with threshold filtering to remove noise spikes, or wavelet transform modulus maxima can be used to locate the true abrupt change point. The analysis time can be the timestamp corresponding to the maximum point, used to focus on the analysis of voltage abrupt events. In a specific embodiment, the analysis time can be directly taken as the timestamp of the sampling time where the maximum point is located. Furthermore, the analysis time can discretize the continuous disturbance signal into key event points, reducing computational complexity and improving feature representativeness. For example, the analysis time can be stored as a sampling sequence number as a relative time, or converted to absolute system time for cross-cycle alignment.

[0062] The slope of the input voltage deviation can be the rate of change of the input voltage deviation over time, usually approximated by the difference between adjacent sampling points. It can be used to characterize the dynamic severity of voltage disturbances and to identify the start and end of abrupt changes. In this embodiment, the slope of the input voltage deviation is obtained by performing a first-order difference on the input voltage deviation sequence or by differentiating using a Savitzky-Golay filter. Furthermore, the slope of the input voltage deviation can be improved by using center difference, or by using low-pass filtering followed by difference to suppress the influence of high-frequency noise. The local voltage abrupt change degree can be the average of the absolute differences between the input voltage deviation slopes at a certain analysis time and the other times in its neighborhood. It can be used to quantify the degree of anomalousness of the voltage change rate at that analysis time relative to the local background. In a specific embodiment, the local voltage abrupt change degree is obtained by calculating the average of the absolute values ​​of the differences between the slope at the analysis time and the slopes at each neighboring time. Furthermore, the local voltage abrupt change degree can be improved by using a weighted average to highlight the influence of neighboring times, or by using the median instead of the mean to improve noise resistance.

[0063] The overall voltage fluctuation degree can be the arithmetic mean of the local voltage fluctuation degrees at all analysis times, and can be used to reflect the average severity of voltage fluctuation events over the entire historical analysis period. In this embodiment, the overall voltage fluctuation degree is obtained by summing all local voltage fluctuation degrees and dividing by the total number of analysis times. Further, the overall voltage fluctuation degree can be normalized by introducing the number of valid analysis times, or by removing local fluctuation degrees below a threshold before averaging. Two adjacent analysis times can be time pairs consisting of two consecutive analysis times in chronological order, and can be used to measure the amplitude jump characteristics between fluctuation events. In an exemplary embodiment, two adjacent analysis times are obtained by pairing the sorted list of analysis times one by one. Voltage fluctuation degree can be the average of the absolute differences in incoming voltage deviations between all adjacent analysis times, and can be used to characterize the discontinuity and jump amplitude between disturbance events, reflecting the severity of grid state switching. In this embodiment, the voltage fluctuation degree is obtained by averaging the absolute values ​​of the differences in voltage deviations corresponding to each pair of adjacent analysis times. Further, the voltage fluctuation degree can consider only adjacent pairs with time intervals less than the upper limit to exclude sparse event interference, or a geometric mean can be used instead of the arithmetic mean. For example, voltage fluctuations may include, but are not limited to, one or more of short-interval fluctuations, long-interval fluctuations, and mixed-interval fluctuations.

[0064] Obtaining the maximum value of the incoming voltage deviation across all times within the historical analysis period for each moment can be achieved by detecting local maxima locations within the incoming voltage deviation sequence covered by the historical analysis period. Further, this operation can be implemented using peak detection algorithms combined with threshold filtering to remove noise spikes, or by using wavelet transform modulus maxima to locate true abrupt change points. This allows for the extraction of the most significant abrupt change events in voltage disturbances, focusing on key disturbance moments. The time corresponding to the maximum value is recorded as the analysis time, which can be achieved by mapping the detected maximum value index to its corresponding timestamp, forming a list of analysis times. Further, this operation can be implemented by directly storing the sampling sequence number as a relative time, or by converting it to absolute system time for easy cross-cycle alignment, thereby establishing a discretized disturbance event time set to support subsequent abrupt change feature calculations.

[0065] The slope of the incoming voltage deviation at each time step is calculated by performing differential or numerical differentiation operations on the incoming voltage deviation sequence to obtain the rate of change at each time step. Furthermore, this operation can use center difference to improve accuracy, or employ low-pass filtering followed by differential calculation to suppress the influence of high-frequency noise, thereby obtaining dynamic gradient information of the voltage disturbance and providing a basis for calculating the abrupt change degree. The local voltage abrupt change degree at each analysis time is obtained by averaging the absolute differences between the slopes of the incoming voltage deviations at each analysis time and the slopes at other times within its neighborhood time period. This can be achieved by calculating the absolute value of the difference between the slope of each analysis time and the slopes of other times within its neighborhood time period, and then averaging them. Furthermore, this operation can be achieved by using a weighted average to highlight the influence of neighboring times, or by using the median instead of the mean to improve noise resistance, thereby measuring the anomalous intensity of the abrupt event relative to the local background and avoiding misclassification of stable segments as abrupt changes. The average of the local voltage abrupt changes at all analysis times is denoted as the overall voltage abrupt change degree, which can be achieved by calculating the arithmetic mean of the local voltage abrupt changes at all analysis times. Furthermore, this operation can be achieved by introducing a normalization of the number of effective analysis moments, or by removing local mutation degrees below a threshold and then averaging them, thereby generating a global index characterizing the severity of mutations throughout the entire historical period.

[0066] The voltage fluctuation is obtained by averaging the absolute difference of the incoming voltage deviation between any two adjacent analysis times. This can be achieved by iterating through the sorted list of analysis times, calculating the absolute value of the difference in voltage deviation between adjacent times, and then averaging the results. Furthermore, this operation can consider only adjacent pairs with time intervals less than an upper limit to exclude sparse event interference, or by using a geometric mean instead of an arithmetic mean. This allows for the quantification of amplitude jump characteristics between disturbance events, reflecting the discontinuity of grid state switching. Based on the overall voltage abrupt change and voltage fluctuation, the voltage fluctuation chaos coefficient at each time point is obtained. This can be achieved by fusing the overall voltage abrupt change and voltage fluctuation into a single coefficient using a preset function (such as product, weighted sum, or nonlinear mapping). For example, the formula for calculating the voltage fluctuation chaos coefficient can be as follows:

[0067] In the formula, The voltage fluctuation chaos coefficient is used to quantify the degree of disorder in the fluctuation pattern of the incoming line voltage. The larger the value, the more unstable the grid voltage is and the more difficult it is to predict. The overall voltage abrupt change is represented by the mean of the local voltage abrupt change at all analysis times (extreme points), reflecting the steepness of the voltage change. This represents the voltage fluctuation, the average difference between adjacent extreme points, reflecting the amplitude span of voltage changes. Furthermore, this operation can be implemented using table lookup interpolation to achieve two-dimensional mapping, or by constructing a lightweight neural network for nonlinear fusion. This allows for the generation of a chaotic index that comprehensively reflects the severity and discontinuity of voltage disturbances, accurately characterizing the interference features of aging power grids.

[0068] For example, in a scenario where concentrated use of kitchen appliances in an older residential area in the evening causes complex disturbances, the dynamic voltage regulation charging control method for the electric two-wheeler charger in this embodiment can be as follows: Between 18:20 and 18:30, multiple households simultaneously use high-power appliances such as rice cookers and microwave ovens, causing frequent drops in the incoming line voltage accompanied by recovery surges. The charger calculates the voltage fluctuation chaos coefficient at t=18:30: First, it identifies five maximum points (corresponding to surge peaks) from the incoming line voltage deviation over the past 10 minutes and marks them as analysis times; it calculates the voltage slope at each time point and finds that the slopes at two times are much higher than the neighborhood average, indicating high local voltage abruptness; therefore, the overall voltage abruptness is high; at the same time, the voltage deviation difference between adjacent analysis times is large (such as a sudden jump from -15V to +10V), and the voltage fluctuation is also high; the two are combined to generate a high voltage fluctuation chaos coefficient, indicating that the current power grid is in a state of high chaotic disturbance. The input error accumulation factor calculation module prompts the voltage regulation compensation strategy to enhance the differential action to suppress sudden shocks, while limiting integral accumulation to prevent over-adjustment, thereby maintaining stable battery terminal voltage.

[0069] In one embodiment, obtaining the interference voltage fluctuation mapping value at each moment includes: Cable interference index and charging ripple amplitude are recorded as analysis indicators; Curve fitting is performed on the same analytical index for all times within the historical analysis period for each time point to obtain two envelope lines of the fitted curve; phase extraction is performed on the envelope lines to obtain the phase value; the absolute difference between the phase values ​​of the two envelope lines of the fitted curve corresponding to each analytical index is recorded as the phase difference at each time point. Calculate the variance of each analytical indicator across all times within the historical analysis period for each time point, as the data volatility at each time point; Based on the phase difference and data volatility of all types of analytical indicators, the comprehensive interference level at each moment is obtained; Based on the voltage fluctuation chaos coefficient and the overall interference degree, the interference voltage fluctuation mapping value at each moment is obtained.

[0070] The analysis indicators can be a unified set of input variables for interference analysis, consisting of cable interference indicators and charging ripple amplitude, which can serve as the basic data source for subsequent curve fitting and feature extraction. In this embodiment, the analysis indicators can directly reuse the collected cable interference indicators and charging ripple amplitude, grouped by type to form two categories of analysis indicators. For example, the analysis indicators can be organized into two categories in the form of a structure, or identifiers can be used to distinguish analysis indicators from different sources. Curve fitting can be a process of approximating the time series of a certain type of analysis indicator within a historical analysis period, which can be used to smooth the original noise data and extract the main trend and modulation envelope of the interference signal. In an exemplary embodiment, curve fitting can use methods such as least squares, spline interpolation, or low-pass filtering to fit the time series. Furthermore, the fitted curve can be a continuous function expression obtained after curve fitting of a certain type of analysis indicator within a historical analysis period, which can be used as the basic model for extracting the upper and lower envelopes. In a specific embodiment, the fitted curve consists of the analytical or discrete representation output by the curve fitting algorithm.

[0071] The two envelopes can be two time-varying curves representing the upper peak boundary and lower valley boundary of the fitted curve, respectively, and can be used to characterize the dynamic amplitude range and modulation asymmetry of the interference signal. In this embodiment, the upper and lower envelopes can be generated by Hilbert transform, local extremum interpolation, or moving extremum window method. Phase extraction can be an operation to parse the time phase information of the dominant oscillation component from the envelope, which can be used to quantify the temporal characteristics of envelope changes and to detect asymmetric disturbances. In an exemplary embodiment, phase extraction can be performed on the envelope by Fourier transform to obtain the dominant frequency phase, or the instantaneous phase can be extracted by analytical signal method. The phase value can be the time phase angle of the envelope at the dominant frequency obtained after phase extraction, which can be used to characterize the start time and periodicity of envelope changes. In a specific embodiment, the phase value is the phase angle expressed in radians or normalized periodic units.

[0072] Phase difference can be the absolute difference between the phase values ​​of the upper and lower envelopes corresponding to the same analytical indicator, and can be used to reflect the asymmetry and mismatch of the interference signal in the positive and negative directions. In this embodiment, the phase difference is obtained by calculating the absolute value of the difference between the upper and lower envelope phase values. Data volatility can be the variance of the values ​​of a certain analytical indicator at all times within the historical analysis period, and can be used to measure the amplitude dispersion of the interference indicator, reflecting its instability. In an exemplary embodiment, data volatility is obtained by calculating the sample variance of the indicator sequence within the historical period. Furthermore, data volatility can use the unbiased sample variance formula, or a moving variance recursive algorithm can be used to reduce computational overhead. Comprehensive interference intensity can be a multi-dimensional comprehensive interference intensity index generated by fusing the phase differences and data volatility of all analytical indicators, and can be used to uniformly characterize the comprehensive impact of interference in both morphological complexity and amplitude volatility dimensions. In a specific embodiment, comprehensive interference intensity is generated by weighted summation, normalized multiplication, or fuzzy fusion of the phase differences and data volatility of each indicator.

[0073] Cable interference indicators and charging ripple amplitude are denoted as analysis indicators, which can classify two types of collected interference-related signals into a unified analysis input. Furthermore, this operation can be achieved by organizing the two types of indicators in a structured form or by using identifiers to distinguish analysis indicators from different sources, thus establishing a structured multi-source interference analysis framework. Curve fitting is performed on the same analysis indicator for all times within the historical analysis period at each moment, obtaining two envelopes of the fitted curve. This can be done by independently fitting the sequence of each type of analysis indicator within the historical period and generating upper and lower envelopes based on the fitting results. Further, this operation can be achieved by using cubic spline fitting followed by local maxima / minimum interpolation to obtain the envelope, or by using Empirical Mode Decomposition (EMD) combined with Hilbert transform to generate the envelope, thereby separating the trend and boundary of the interference signal and supporting morphological feature extraction.

[0074] Phase extraction is performed on the envelope to obtain phase values. This can be achieved by performing frequency domain or time-frequency analysis on the upper and lower envelopes respectively to extract the phase of their dominant components. Further, this operation can be implemented by performing an FFT on the envelope to obtain the dominant frequency phase, or by constructing an analytic signal and calculating the instantaneous phase using the atan2 function, thus obtaining a time reference for envelope changes to detect asymmetric disturbances. The absolute difference between the phase values ​​of the two envelopes corresponding to the fitted curve for each analysis index is recorded as the phase difference at each moment. This can be achieved by calculating the absolute difference between the upper and lower envelope phase values ​​of the same index to form the phase difference for that index. Further, this operation can be implemented by directly calculating the phase angle difference and taking its absolute value, or by considering the phase wrapping problem and using a minimum angle difference algorithm, thus quantifying the temporal asymmetry of the positive and negative half-cycles of the interference signal.

[0075] Calculate the variance of each analytical indicator across all times within the historical analysis period for each moment, using this variance as the data volatility at each moment. This can be achieved by calculating the variance of each analytical indicator's sequence over the historical period. Furthermore, this operation can be implemented by using the unbiased sample variance formula or a moving variance recursive algorithm to reduce computational overhead, thus reflecting the randomness and severity of the interference source's amplitude. Obtain the comprehensive interference intensity at each moment based on the phase difference and data volatility of all types of analytical indicators. This can be achieved by multi-dimensionally fusing the phase difference and data volatility of each of the two types of analytical indicators. Further, this operation can be achieved by normalizing and weighting the four features (two types of indicators × phase difference / volatility), or by inputting them into a lightweight classifier to output a comprehensive score, thereby generating a unified interference intensity index that considers both morphological complexity and amplitude instability. Obtain the interference voltage fluctuation mapping value at each moment based on the voltage fluctuation chaos coefficient and the comprehensive interference intensity. This can be achieved by fusing the voltage fluctuation chaos coefficient (characterizing grid abrupt changes) and the comprehensive interference intensity (characterizing local interference complexity) into a final mapping value. The formula for calculating the interference voltage fluctuation mapping value can be as follows:

[0076] In the formula, The comprehensive interference level refers to the sum of cable interference and ripple interference faced by the charger at time t. or Phase difference refers to the absolute difference between the upper and lower envelopes of the fitted curve obtained after curve fitting of charging ripple amplitude or cable interference index data within a historical analysis period. The envelope phase difference reflects the asymmetry or distortion degree of the waveform. or Data volatility / variance refers to the variance of charging ripple amplitude or cable interference index data at all times during the historical analysis period. Variance reflects the degree of dispersion of the data in amplitude, that is, the degree of drastic fluctuation of the value. The voltage fluctuation chaos coefficient; This is a mapping value for interference voltage fluctuations, mapping environmental interference to specific impact values ​​on voltage control for subsequent compensation calculations. Furthermore, this operation can emphasize the superposition effect of high chaos and high interference by using a product form, or achieve nonlinear mapping through a two-dimensional lookup table, thereby forming a full-link interference characterization covering external power grid disturbances and local line interference.

[0077] For example, in a scenario of multiple sources of interference superimposed in an old community in the evening, the dynamic voltage regulation charging control method of the electric two-wheeler charger in this embodiment can be implemented during the peak electricity consumption period of 19:00, when a nearby shop starts a high-power freezer causing a surge, and at the same time, the user's use of a poor-quality extension cord causes high-frequency noise coupling. The charger calculates the interference voltage fluctuation mapping value at t=19:05: First, the cable interference index (high) and the charging ripple amplitude (significantly increased) are used as two types of analysis indicators; after fitting their respective historical 5-minute data, it is found that the upper and lower envelope phase difference of the ripple amplitude is large (due to the surge causing a prominent positive spike), while the envelope phase difference of the cable interference index is small but the variance is extremely high (due to continuous high-frequency noise); therefore, the overall interference level is high; then, combined with the previously calculated high voltage fluctuation chaos coefficient (due to voltage abrupt changes caused by the freezer starting and stopping), the two are fused to generate a high interference voltage fluctuation mapping value. This value prompts the voltage regulation compensation coefficient to enhance feedforward suppression and PID differential gain, effectively suppressing ripple propagation and accelerating the response to surges, thus keeping the battery terminal voltage stable.

[0078] In one embodiment, obtaining the voltage regulation compensation coefficient at the current moment includes: The interference voltage fluctuation mapping values ​​and voltage regulation response hysteresis factors of all times in the historical analysis period of the current time are arranged in time sequence to obtain the fluctuation mapping sequence and hysteresis factor sequence. The correlation coefficient between the two sequences is obtained and recorded as the compensation confidence level at the current time. The product of the current disturbance voltage fluctuation mapping value and the voltage regulation response hysteresis factor is used as the disturbance severity. The voltage regulation compensation coefficient at the current moment is obtained based on the severity of the interference, the compensation confidence level, and the error accumulation factor.

[0079] The fluctuation mapping sequence can be a time-series data sequence formed by arranging the interference voltage fluctuation mapping values ​​at each moment within the historical analysis period of the current moment in chronological order. It can be used to characterize the evolution characteristics of external interference in the time dimension and support correlation analysis. In this embodiment, the fluctuation mapping sequence can be extracted from the calculated interference voltage fluctuation mapping values ​​at each moment and organized into a one-dimensional array according to the sampling time sequence. The hysteresis factor sequence can be a time-series data sequence formed by arranging the voltage regulation response hysteresis factors at each moment within the historical analysis period of the current moment in chronological order. It can be used to characterize the change law of the system's dynamic response capability over time and support coupled analysis with the interference sequence. For example, the hysteresis factor sequence can be extracted from the calculated voltage regulation response hysteresis factors at each moment and organized into a one-dimensional array according to the sampling time sequence.

[0080] The correlation coefficient can be a statistic measuring the degree of linear correlation between the fluctuation mapping sequence and the hysteresis factor sequence. It can be used to quantify the synchronicity or causal relationship between the intensity of grid disturbances and the system response hysteresis. In an exemplary embodiment, the correlation coefficient is calculated using the Pearson correlation coefficient formula for two time series of equal length. The compensation confidence level can be a dynamic indicator reflecting the reliability of the voltage regulation compensation strategy, defined by the correlation coefficient between the fluctuation mapping sequence and the hysteresis factor sequence. It can be used as a weighting factor in the generation process of the voltage regulation compensation coefficient to adjust the compensation intensity to avoid over-response in a highly random environment. Furthermore, the compensation confidence level is directly assigned from the correlation coefficient and participates in the final synthesis of the voltage regulation compensation coefficient, working together with the disturbance severity and error accumulation factor. The disturbance severity can be the product of the disturbance voltage fluctuation mapping value at the current moment and the voltage regulation response hysteresis factor, characterizing the degree of joint impact of disturbance intensity and system vulnerability. It can be used to comprehensively reflect the risk level faced by the charging system under the current operating conditions, avoiding misjudgments caused by relying solely on a single disturbance or response indicator. In one specific embodiment, the interference severity is obtained by multiplying the currently calculated interference voltage fluctuation mapping value by the voltage regulation response hysteresis factor. Furthermore, the interference severity, as one of the key inputs to the voltage regulation compensation coefficient, together with the compensation confidence level and the error accumulation factor, determines the final compensation strength.

[0081] The disturbance voltage fluctuation mapping values ​​and voltage regulation response hysteresis factors for all moments within the historical analysis period at the current moment are arranged chronologically to obtain the fluctuation mapping sequence and hysteresis factor sequence. This can be achieved by extracting the two types of dynamic factors within a specified time window from the historical buffer and organizing them into two independent one-dimensional sequences according to the sampling timestamp order. Furthermore, this operation can be implemented by using a circular buffer to store historical factor values ​​and directly outputting an ordered sequence, or by querying and sorting sequences from a database using timestamp indexes, thus providing a structured time-series data foundation for subsequent correlation analysis. The correlation coefficient between the fluctuation mapping sequence and the hysteresis factor sequence is obtained and denoted as the compensation confidence level at the current moment. This can be achieved by applying a statistical correlation algorithm to two equally time-length sequences to calculate their Pearson correlation coefficient and using the result as the compensation confidence level. Further, this operation can be achieved by calculating the correlation coefficient using the covariance and standard deviation within a sliding window, or by using a fast Fourier transform to estimate phase consistency in the frequency domain to approximate the correlation, thereby quantifying the coupling regularity between disturbance and response and improving the credibility assessment capability of compensation decisions.

[0082] The product of the current disturbance voltage fluctuation mapping value and the voltage regulation response hysteresis factor is used as the disturbance severity. This can be achieved by multiplying the two scalar factors obtained at the current moment to generate a joint risk index. Furthermore, this operation can be achieved by directly multiplying to generate the original severity, or by introducing a nonlinear function (such as logarithmic compression or saturation limiting) before multiplication to suppress the influence of extreme values. This allows for the coupled quantification of disturbance strength and system response vulnerability, enhancing the ability to identify complex operating conditions. Based on the disturbance severity, compensation confidence, and error accumulation factor, the voltage regulation compensation coefficient at the current moment is obtained. This can be achieved by using the three dynamic indicators as input variables and generating the final voltage regulation compensation coefficient through weighted fusion, table lookup mapping, or function fitting. The formula for calculating the voltage regulation compensation coefficient can be as follows:

[0083] In the formula, The voltage regulation compensation coefficient is a dynamic adjustment factor calculated at the current time t. It is a dimensionless coefficient used to adjust the gain parameter of the PID controller in real time, and its magnitude determines the correction strength. The adjustment coefficient / proportional constant is a preset empirical constant or proportional factor used to match the system dimensions, preventing the calculated compensation coefficient from being too large, causing the system to over-adjust, or too small, causing insufficient compensation. The interference voltage fluctuation mapping value quantifies environmental factors such as cable interference and ripple and maps them to a numerical value that affects voltage fluctuations, reflecting how harsh the external environment is. The voltage regulation response hysteresis factor is the degree to which the actual voltage of the charger follows the theoretical voltage over a historical period, reflecting how slow the system's response is. The severity of the interference is the product of the interference mapping value and the hysteresis factor. To compensate for the confidence level, the correlation coefficient between the historical disturbance voltage fluctuation mapping value sequence and the voltage regulation response hysteresis factor sequence, such as the Pearson correlation coefficient, is a value between 0 and 1, reflecting the closeness of the correlation between external disturbance fluctuations and system hysteresis behavior. The error accumulation factor is a comprehensive index combining the degree of chaos in the incoming line voltage, response hysteresis, and the intensity of error accumulation. It reflects the risk level of continuous accumulation and deterioration of the current error. Furthermore, this operation can be achieved by using a linearly weighted combination followed by Sigmoid function normalization, or by constructing a lightweight multi-input single-output neural network model for online inference of the compensation coefficients. This allows the compensation coefficients to simultaneously include the current risk level, the reliability of historical behavior patterns, and the state of deviation accumulation, improving the comprehensiveness and robustness of adaptive adjustment.

[0084] For example, in a scenario where electricity consumption fluctuates drastically in an older community during the evening, the dynamic voltage regulation charging control method for the electric two-wheeler charger in this embodiment could be as follows: In the evening, residents concentrate on turning on high-power appliances, causing frequent voltage drops and surges in the incoming line. The charger detects a continuous increase in the interference voltage fluctuation mapping value, and simultaneously, due to the long cable, the voltage regulation response hysteresis factor increases synchronously. The system constructs a fluctuation mapping sequence and a hysteresis factor sequence from the two types of factors over the past 30 seconds, calculating a correlation coefficient of 0.82, denoted as high compensation confidence. The product of the two at the current moment forms a high interference severity, which, combined with the medium error accumulation factor accumulated due to previous undervoltage, ultimately generates a large voltage regulation compensation coefficient. This coefficient appropriately increases the PID proportional gain and raises the target voltage, effectively compensating for line voltage drop while ensuring no overshoot, allowing the battery terminal voltage to stably track the constant current curve and avoiding repeated start-stop cycles.

[0085] In one embodiment, dynamic voltage regulation charging control of the electric two-wheeler charger is performed using a voltage regulation compensation coefficient, including: Obtain the ideal PID voltage regulation parameters of the charger, and use the sum of constant 1 and the voltage regulation compensation coefficient at the current moment to weight the ideal PID voltage regulation parameters to obtain the corrected PID voltage regulation parameters at the current moment; use the corrected PID voltage regulation parameters to perform real-time dynamic voltage regulation charging control of the electric two-wheeler charger.

[0086] The ideal PID voltage regulation parameters can be the PID controller reference parameters that enable the charger output voltage to optimally track the ideal charging curve under standard power grid conditions and a disturbance-free environment. These parameters can be used as a reference value for dynamic correction, ensuring that parameter adjustments have a reasonable starting point. In this embodiment, the ideal PID voltage regulation parameters can be obtained through offline simulation, experimental calibration, or theoretical design and stored in the controller firmware. The constant 1 can be a fixed scalar with a value of 1, which can be added to the voltage regulation compensation coefficient to form the basic bias term of the adaptive gain factor, ensuring that the corrected parameters are always scaled based on the original parameters. In an exemplary embodiment, the constant 1 is directly used as a constant in the algorithm. The sum can be the scalar result obtained by adding the constant 1 to the voltage regulation compensation coefficient at the current moment, which can be used as a dynamic gain factor for proportionally weighting the ideal PID voltage regulation parameters. Furthermore, the sum is calculated in real time using arithmetic addition.

[0087] The corrected PID voltage regulation parameters can be real-time PID control parameters suitable for the current power grid disturbance state, obtained by multiplying the ideal PID voltage regulation parameters by a sum. These parameters can be used to drive the charger to perform dynamic voltage regulation behavior matched to the current environment, improving response accuracy and stability. In one embodiment, the corrected PID voltage regulation parameters are scaled overall by multiplying each term of the ideal PID voltage regulation parameters by a sum. For example, the corrected PID voltage regulation parameters are generated based on the voltage regulation compensation coefficient and the ideal PID voltage regulation parameters, and are used for real-time dynamic voltage regulation charging control. Obtaining the charger's ideal PID voltage regulation parameters can be achieved by reading preset reference PID parameters from the controller's storage unit. Furthermore, obtaining the charger's ideal PID voltage regulation parameters can be achieved by loading pre-stored parameters from non-volatile memory or by automatically selecting a corresponding set of ideal PID parameters based on the battery type, thereby providing a stable and reliable reference benchmark for subsequent dynamic corrections.

[0088] The ideal PID voltage regulation parameters are weighted using the sum of a constant 1 and the current voltage regulation compensation coefficient. This can be achieved by multiplying each term of the ideal PID voltage regulation parameters by (1 + voltage regulation compensation coefficient) and summing the results. Furthermore, this operation can be achieved by multiplying the P, I, and D terms by the same sum, or by applying different weights to different terms and then combining them with the sum for differentiated scaling. This allows the control strength to adapt to the degree of disturbance. The corrected PID voltage regulation parameters for the current moment can be obtained by outputting a set of PID parameters usable in the current control cycle after weighted calculation. The formula for calculating the corrected PID voltage regulation parameters is as follows:

[0089] In the formula, To correct the PID voltage regulation parameters, the actual PID parameters used by the charger during operation; These are the optimal PID parameters set under standard power grid conditions, representing the ideal PID voltage regulation parameters. As a weighting factor, based on the calculated voltage regulation compensation coefficient By changing the gain of the reference parameter, adaptive control can be achieved for the complex power grid environment of older communities. Furthermore, the corrected PID voltage regulation parameters for the current moment can be obtained by directly replacing the contents of the original PID parameter register, or by writing the corrected parameters into a dedicated buffer for use in the next control cycle, thus generating real-time control parameters adapted to the current power grid state. Real-time dynamic voltage regulation charging control of the electric two-wheeler charger using the corrected PID voltage regulation parameters can be achieved by applying the corrected PID parameters to perform closed-loop voltage regulation within the current control cycle. Furthermore, this operation can be achieved by inputting the corrected parameters into the digital PID controller hardware module, or by dynamically updating the PID algorithm coefficients in the software control loop, thereby enabling the output voltage to respond quickly and accurately to disturbance changes and maintain stable battery terminal voltage.

[0090] For example, in a scenario where peak electricity consumption coincides with thunderstorms in an older community, the dynamic voltage regulation charging control method for the electric two-wheeler charger in this embodiment can be as follows: The grid voltage drops to 185V due to a surge in load, while a high-frequency surge is induced by lightning. The charger detects a significant input voltage deviation and high ripple, and calculates a voltage regulation compensation coefficient of 0.35 based on previous steps. The system reads the ideal PID voltage regulation parameters (Kp = 2.0, Ki = 0.5, Kd = 0.1), calculates the sum as 1 + 0.35 = 1.35, and multiplies each parameter by this value to obtain the corrected PID voltage regulation parameters (Kp = 2.7, Ki = 0.675, Kd = 0.135). The enhanced proportional and integral gains accelerate the compensation speed for voltage drops, allowing the battery terminal voltage to recover to the constant current stage set value within 300ms, avoiding triggering undervoltage protection shutdown and ensuring charging continuity.

[0091] In one embodiment, the overall interference level at each time step is obtained, including: Calculate the product of the phase difference and data volatility for each analytical indicator at each time step, and use it as the local interference degree; sum the local interference degrees of all types of analytical indicators at each time step, and use it as the comprehensive interference degree at each time step.

[0092] The local interference degree can be the product of the phase difference and data volatility of a single type of analytical indicator at a certain moment. It can be used to jointly characterize the coupling strength of the interference source in two dimensions: morphological asymmetry and amplitude instability. In this embodiment, the local interference degree can be obtained by multiplying the phase difference and data volatility corresponding to the same analytical indicator. Calculating the product of the phase difference and data volatility corresponding to each analytical indicator at each moment as the local interference degree can be achieved by multiplying the phase difference and data volatility of each analytical indicator at the current moment to generate the local interference degree of that indicator. Furthermore, this operation can be implemented by directly performing floating-point multiplication; or by approximating the product through shifting and table lookup in a fixed-point system to reduce resource consumption, thereby achieving nonlinear coupling between morphological complexity and amplitude volatility, highlighting the comprehensive impact of high mutation and high discrete interference. The sum of the local interference degrees of all types of analytical indicators at each moment can be used as the comprehensive interference degree at each moment. This can be achieved by summing the local interference degrees of all analytical indicators (such as cable interference indicators and charging ripple amplitude). In one specific embodiment, this operation can be achieved by using an unsigned accumulator to accumulate item by item; or by introducing weighting coefficients to perform weighted summation to distinguish the importance of different interference sources, thereby integrating the coupling effect of multi-source interference and generating a unified comprehensive interference intensity scalar.

[0093] For example, in a scenario where multiple devices start and stop at night in an old community coexist with substandard cables, the dynamic voltage regulation charging control method for electric two-wheeled vehicle chargers in this embodiment could be as follows: At 21:00, a user uses an aging extension cord to charge their electric vehicle, while upstairs residents frequently turn their air conditioners on and off. The charger collects cable interference indicators with high variance but small phase difference (continuous high-frequency noise), while the charging ripple amplitude experiences significant surges due to the air conditioner's start and stop, exhibiting high phase difference and high volatility. After calculating the local interference degree for both, the ripple term product is extremely large, while the cable term product is moderate; summing these results in a high overall interference degree. This value is input into the interference voltage fluctuation mapping value generation module, combined with the high voltage fluctuation chaos coefficient, ultimately triggering a strong compensation strategy: increasing the feedforward gain and enhancing the PID differential action, effectively suppressing ripple transmission to the battery terminal, and maintaining voltage stability during the constant current phase.

[0094] In one embodiment, the least squares method is used to perform curve fitting on the same analytical index for all times within the historical analysis period at each time point.

[0095] The least squares method can be a mathematical optimization method that estimates model parameters by minimizing the sum of squared residuals between observed data and model predictions. It can be used to smooth the time series of noisy analytical indicators and extract their inherent trends. In this embodiment, the least squares method can define basis functions (such as polynomials or exponential functions) and solve for the coefficient vector that minimizes the sum of squared residuals. Curve fitting of the same analytical indicator at all times within the historical analysis period for each time point can be performed using the least squares method. This can be done for each type of analytical indicator (such as cable interference or charging ripple amplitude), within the time window covered by its historical analysis period, with time as the independent variable and the indicator value as the dependent variable. Furthermore, this operation can be achieved by using low-order polynomials (such as quadratic or cubic) for global fitting to avoid overfitting; or by using sliding window local least squares fitting to adapt to non-stationary characteristics. This allows for the extraction of smooth and faithful trend curves from the original data containing high-frequency noise and random disturbances, providing a stable foundation for subsequent envelope and phase analysis.

[0096] For example, in a scenario where nighttime charging in an older community is accompanied by intermittent starting and stopping of high-power appliances, the dynamic voltage regulation charging control method for the electric two-wheeler charger in this embodiment can be as follows: After 21:00, multiple households' air conditioners operate intermittently, causing periodic spikes and random jumps in the charging ripple amplitude sequence. At t=21:10, the charger applies the least squares method to perform a cubic polynomial fitting on the charging ripple amplitude sequence of the past 5 minutes, effectively filtering out instantaneous switching noise and retaining the continuous high ripple trend caused by voltage drops. This fitted curve is then used to generate upper and lower envelopes, whose phase characteristics clearly reflect the dominant frequency of the disturbance, avoiding false phase jumps caused by isolated spikes in the original data, thereby ensuring that the interference voltage fluctuation mapping value accurately represents the real power grid interference intensity.

[0097] In addition, to achieve the above objectives, the present invention also provides a computer-readable storage medium storing a dynamic voltage regulation charging control program for an electric two-wheeled vehicle charger. When the dynamic voltage regulation charging control program for an electric two-wheeled vehicle charger is executed by a processor, it implements the steps of the dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger as described above.

[0098] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger, characterized in that, The method includes: Real-time acquisition of the incoming voltage deviation, cable interference index, charging ripple amplitude, actual battery terminal charging voltage, and theoretical battery terminal charging voltage under voltage regulation control command of the electric two-wheeler charger at each moment; Based on the deviation and fluctuation of the actual battery terminal charging voltage and the theoretical battery terminal charging voltage at each time point in the historical analysis period, the voltage regulation response hysteresis factor at each time point is obtained; based on the voltage regulation response hysteresis factor at the current time point and the degree of change of the incoming line voltage deviation at each time point in the historical analysis period, the error accumulation factor at the current time point is obtained. Based on the historical analysis of cable interference indicators, the degree of disorder and fluctuation of charging ripple amplitude, and the degree of change of incoming line voltage deviation at each time point within the historical analysis period, the interference voltage fluctuation mapping value at each time point is obtained. Based on the error accumulation factor, the interference voltage fluctuation mapping value, the voltage regulation response hysteresis factor at the current moment, and the correlation between the interference voltage fluctuation mapping value and the voltage regulation response hysteresis factor at each moment in the historical analysis period at the current moment, the voltage regulation compensation coefficient at the current moment is obtained; the voltage regulation compensation coefficient is used to perform dynamic voltage regulation charging control on the electric two-wheeler charger.

2. The dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger as described in claim 1, characterized in that, The process of obtaining the voltage regulation response hysteresis factor at each moment includes: The theoretical battery terminal charging voltage and the actual battery terminal charging voltage at all times within the historical analysis period of each time are arranged in time sequence to obtain the theoretical voltage sequence and the actual voltage sequence. The theoretical voltage sequence and the actual voltage sequence are time-series matched. The absolute value of the deviation between the actual battery terminal charging voltage in the actual voltage sequence and the theoretical battery terminal charging voltage matched in the theoretical voltage sequence is calculated. The absolute values ​​of the deviations corresponding to all actual battery terminal charging voltages in the actual voltage sequence are arranged to obtain the voltage error sequence at each time point. Calculate the variance of all elements within a preset sliding window for each element in the voltage error sequence, and denote it as the neighborhood deviation volatility of each element; arrange the neighborhood deviation volatility of all elements in the voltage error sequence to obtain the deviation volatility sequence. The deviation fluctuation sequence is integrated over the historical analysis period at each moment, and the ratio of the integration result to the duration of the historical analysis period is used as the voltage regulation response hysteresis factor at each moment.

3. The dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger as described in claim 2, characterized in that, The step of obtaining the error accumulation factor at the current moment includes: Based on the differences in the rate of change of incoming line voltage deviation between each time period and its neighboring time periods in the historical analysis period, as well as the differences in incoming line voltage deviation between adjacent time periods, the voltage fluctuation chaos coefficient at each time period is obtained. Obtain the first-order difference sequence of the voltage error sequence at the current moment, and use the mean of the elements in the first-order difference sequence as the cumulative error intensity at the current moment. The error accumulation factor at the current moment is obtained based on the voltage fluctuation chaos coefficient, the voltage regulation response hysteresis factor, and the error accumulation intensity.

4. The dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger as described in claim 3, characterized in that, The acquisition of the voltage fluctuation chaos coefficient at each moment includes: Obtain the maximum value point of the incoming line voltage deviation at all times within the historical analysis period for each time moment, and record the time corresponding to the maximum value point as the analysis time moment; Calculate the slope of the incoming line voltage deviation at each time point, and average the absolute difference between the slopes of the incoming line voltage deviation at each analysis time point and the slopes at other times in the neighboring time period to obtain the local voltage mutation degree at each analysis time point; the mean of the local voltage mutation degrees at all analysis times is denoted as the overall voltage mutation degree. The voltage fluctuation is obtained by averaging the absolute difference between the incoming line voltage deviations at each two adjacent analysis times. Based on the overall voltage abrupt change and the voltage fluctuation, the voltage fluctuation chaos coefficient at each moment is obtained.

5. The dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger as described in claim 4, characterized in that, The process of obtaining the interference voltage fluctuation mapping value at each moment includes: Cable interference index and charging ripple amplitude are recorded as analysis indicators; Curve fitting is performed on the same analytical index for all times within the historical analysis period for each time point to obtain two envelope lines of the obtained fitted curve; phase extraction is performed on the envelope lines to obtain phase values; the absolute difference of the phase values ​​of the two envelope lines of the fitted curve corresponding to each analytical index is recorded as the phase difference at each time point. Calculate the variance of each analytical indicator across all times within the historical analysis period for each time point, as the data volatility at each time point; Based on the phase difference and data volatility of all types of analytical indicators, the comprehensive interference degree at each moment is obtained; Based on the voltage fluctuation chaos coefficient and the comprehensive interference degree, the interference voltage fluctuation mapping value at each moment is obtained.

6. The dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger as described in claim 1, characterized in that, The step of obtaining the voltage regulation compensation coefficient at the current moment includes: The interference voltage fluctuation mapping value and the voltage regulation response hysteresis factor of all times in the historical analysis period of the current time are arranged in time sequence to obtain the fluctuation mapping sequence and the hysteresis factor sequence in turn. The correlation coefficient between the two sequences is obtained and recorded as the compensation confidence level at the current time. The product of the current disturbance voltage fluctuation mapping value and the voltage regulation response hysteresis factor is used as the disturbance severity. The voltage regulation compensation coefficient at the current moment is obtained based on the severity of the interference, the compensation confidence level, and the error accumulation factor.

7. The dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger as described in claim 1, characterized in that, The method of using the voltage adjustment compensation coefficient to perform dynamic voltage regulation charging control on the electric two-wheeler charger includes: Obtain the ideal PID voltage regulation parameters of the charger, and use the sum of the constant 1 and the voltage regulation compensation coefficient at the current moment to weight the ideal PID voltage regulation parameters to obtain the corrected PID voltage regulation parameters at the current moment; use the corrected PID voltage regulation parameters to perform real-time dynamic voltage regulation charging control on the electric two-wheeler charger.

8. The dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger as described in claim 5, characterized in that, The process of obtaining the comprehensive interference degree at each moment includes: Calculate the product of the phase difference and the data volatility for each analytical indicator at each time step, and use it as the local interference degree; sum the local interference degrees of all types of analytical indicators at each time step, and use it as the comprehensive interference degree at each time step.

9. The dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger as described in claim 5, characterized in that, The least squares method is used to perform curve fitting on the same analytical index for all times within the historical analysis period at each time point.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a dynamic voltage regulation charging control program for an electric two-wheeled vehicle charger. When the dynamic voltage regulation charging control program for an electric two-wheeled vehicle charger is executed by a processor, it implements the steps of the dynamic voltage regulation charging control method for an electric two-wheeled vehicle charger as described in any one of claims 1 to 9.

Images