A method, medium, and system for automatic adjustment of charging current in high-power charging piles

Abstract

This invention provides a method, medium, and system for automatic adjustment of charging current in high-power charging piles, belonging to the field of charging pile technology. This invention collects key battery parameters in real time, establishes a charging power offset matrix to quantify power deviation, constructs an adjustable current adjustment matrix to define a safe adjustment range, generates a harmonic interference probability matrix to assess the impact of grid interference, establishes an upper-level game model aimed at maximizing charging efficiency and a lower-level game model aimed at optimizing battery safety, uses a time-series adjustment model to process multi-dimensional monitoring data and outputs suggested current adjustment values, and adjusts the neuron activation threshold using appropriate gain coefficients and damping parameters according to different operating conditions through a charging current adaptive function, ultimately achieving intelligent adaptive output of the charging current. This solves the technical problem that charging piles cannot achieve intelligent adaptive adjustment of charging current in complex grid environments.

Description

Technical Field

[0001] This invention belongs to the field of charging pile technology, and specifically relates to a method, medium and system for automatic adjustment of charging current of high-power charging piles. Background Technology

[0002] With the rapid development of the electric vehicle industry, high-power charging piles, as a critical infrastructure, rely on precise control of charging current based on battery status and grid conditions. Traditional charging pile systems primarily employ fixed-parameter PID control algorithms or simple fuzzy control strategies to adjust the charging current. These methods can meet basic requirements under stable grid environments and standard charging conditions, and are widely used in commercial charging stations, residential charging facilities, and highway service areas. However, in today's complex grid environment, due to the coupled effects of multiple factors such as grid voltage fluctuations, harmonic interference, and load abrupt changes, traditional control methods struggle to accurately perceive changes in battery status and grid quality fluctuations, failing to achieve dynamic optimization and adjustment of charging parameters. This results in low charging efficiency and increased safety risks. In practical applications, fixed control strategies often ignore the nonlinear changes in battery charging characteristics and the time-varying characteristics of the grid environment, lacking the ability to deeply model multi-parameter coupling relationships, and thus failing to maximize charging efficiency while ensuring charging safety. In other words, existing technologies suffer from the technical problem of charging piles being unable to achieve intelligent adaptive adjustment of charging current in complex grid environments. Summary of the Invention

[0003] In view of this, the present invention provides a method, medium and system for automatic adjustment of charging current of high-power charging piles, which can solve the technical problem that charging piles cannot achieve intelligent adaptive adjustment of charging current in complex power grid environments in the prior art.

[0004] The present invention is implemented as follows: The first aspect of the present invention provides a method for automatically adjusting the charging current of a high-power charging pile, including constructing a battery state monitoring matrix to establish a basic data structure for battery state assessment by real-time acquisition of parameters such as battery voltage, battery current, battery temperature, and battery internal resistance; establishing a charging power offset matrix to calculate the deviation between the charging pile output power and the target power based on grid voltage fluctuations and load changes, forming a data basis for power offset compensation; constructing an adjustable current adjustment matrix to establish a dynamic adjustment range matrix for the charging current based on the battery charging characteristic curve and safety threshold constraints; generating a harmonic interference probability matrix to detect grid harmonic content using frequency domain analysis methods, calculating the interference probability of different frequency harmonics on the charging process, and establishing a harmonic impact assessment model; establishing an upper-level game model with the goal of maximizing charging efficiency and a lower-level game model with the goal of optimizing battery safety, with the two models achieving coordinated optimization through a charging current coupling function; using a pre-trained liquid-state timing adjustment model to process the battery state monitoring matrix and the charging power offset matrix, outputting optimized charging current adjustment suggestions; and calculating the final charging current output value through a charging current adaptive function.

[0005] The battery status monitoring matrix has rows corresponding to the types of monitoring parameters and columns corresponding to the time series. It is used to store and organize time-series data of key battery status parameters. Each matrix element represents the measured values ​​of battery voltage, battery current, battery temperature, and battery internal resistance at a given time. Multi-parameter fusion analysis is achieved through matrix operations.

[0006] The charging power offset matrix is ​​used to quantify the difference between the actual output power and the ideal output power of the charging pile, providing data support for power adjustment. Each row of the charging power offset matrix corresponds to a different power level, and each column corresponds to a different compensation strategy.

[0007] The adjustable current matrix is ​​used to define the safe adjustment range of the charging current. The elements of the adjustable current matrix represent the charging current adjustment boundary under different operating conditions. The elements of the adjustable current matrix are determined according to the battery type, charging stage and environmental conditions to ensure that the charging process is carried out within the safe boundary.

[0008] Among them, the harmonic interference probability matrix is ​​used to assess the potential impact of power grid harmonics on charging equipment. By statistically analyzing historical harmonic data and real-time monitoring results, the probability of occurrence and the degree of impact of harmonic interference are predicted.

[0009] The objective function of the upper-level game model is to maximize charging efficiency. The inputs include the output power of the charging pile, the suggested value of charging current adjustment, the charging time and the battery voltage. The output is the optimized value of charging efficiency. The constraints include that the charging power does not exceed the rated power of the equipment, the charging current is within the safe range, and the battery temperature is below the safe threshold.

[0010] The objective function of the lower-level game model is the optimal function for battery safety. The inputs include battery temperature, recommended charging current adjustment value, battery internal resistance and voltage difference. The output is the battery safety assessment value. The constraints include battery voltage not exceeding the cutoff voltage, limited charging current change rate, and battery internal resistance change within the normal range.

[0011] The charging current coupling function is used to coordinate the optimization process of the upper-level game model and the lower-level game model. The inputs include the optimized charging efficiency value and the battery safety assessment value, and the output is the coordinated charging current setting reference value.

[0012] The liquid-state time-series conditioning model employs a recurrent neural network architecture based on a liquid state machine, comprising an input layer, a liquid reservoir, a readout layer, and a feedback connection layer. The liquid reservoir consists of numerous randomly connected neurons, with connection weights and time constants dynamically adjusted based on the frequency characteristics of the input signal. The liquid reservoir utilizes a leakage integrator neuron model, exhibiting memory decay and nonlinear activation characteristics. The readout layer extracts liquid reservoir state information through linear combination, while the feedback connection layer feeds back output information to the liquid reservoir to enhance time-series modeling capabilities. The neuron connection weights in the liquid-state time-series conditioning model are dynamically adjusted based on battery temperature, voltage difference, and charging time. Ridge regression is used to train the linear weights of the readout layer, with the optimization objective being to minimize the mean square error between the predicted charging current and the actual optimal charging current.

[0013] The steps for establishing the training dataset for the liquid reservoir time-series regulation model include collecting charging data from different types of batteries under various operating conditions, covering charging modes such as normal charging, fast charging, and trickle charging. Preprocessing of the raw data includes denoising, normalization, and feature extraction. The training steps for the liquid reservoir time-series regulation model include initializing the connection weights and time constants of neurons in the liquid reservoir, training the dynamic characteristics of the liquid reservoir using unsupervised learning methods, stimulating the liquid reservoir to generate rich dynamic response modes by inputting historical charging data sequences, and adjusting the model hyperparameters, including liquid reservoir size, spectral radius, and input scaling factor, using cross-validation.

[0014] The adaptive charging current function calculates the final charging current output value based on the output of the liquid-state timing regulation model and the current system state. Inputs include the recommended charging current regulation value, battery voltage, grid quality indicators, and ambient temperature. The output is the adaptively adjusted charging current output value. When the recommended charging current regulation value is within different ranges, different gain coefficients and damping parameters are used to adjust the neuron activation threshold of the liquid-state timing regulation model. When the recommended charging current regulation value is within the safe charging range, a smaller gain coefficient and moderate damping parameter are used to adjust the neuron activation threshold to maintain a stable output. When the recommended charging current regulation value is close to the safe boundary, a larger gain coefficient and stronger damping parameter are used to increase the neuron activation threshold to enhance the system's conservatism. When the recommended charging current regulation value exceeds the safe range, the maximum gain coefficient and strongest damping parameter are used to significantly increase the neuron activation threshold to ensure system safety.

[0015] A second aspect of the present invention provides a computer-readable storage medium storing program instructions, which, when executed in a computer, are used to perform the above-described method for automatically adjusting the charging current of a high-power charging pile.

[0016] A third aspect of the present invention provides an automatic adjustment system for charging current of a high-power charging pile, comprising the aforementioned computer-readable storage medium. The system is any one of a computer, a server, or a microcontroller. The computer-readable storage medium is disposed within the system, and the system is provided with a microprocessor that executes the program instructions stored in the computer-readable storage medium.

[0017] This invention establishes a multi-dimensional data fusion-based charging state perception system by constructing a battery state monitoring matrix, a charging power offset matrix, an adjustable current regulation matrix, and a harmonic interference probability matrix, achieving comprehensive real-time monitoring of battery state and the power grid environment. By establishing a game-theoretic model between upper and lower layers and a liquid-state time-series adjustment model, this invention constructs a dual-objective coordinated optimization mechanism that maximizes charging efficiency and optimizes battery safety, overcoming the local optimization problem caused by traditional single control objectives and achieving globally optimal configuration of charging parameters. The liquid-state time-series adjustment model, based on a recurrent neural network architecture of a liquid state machine, possesses powerful nonlinear modeling capabilities and time-series memory characteristics, accurately capturing the dynamic changes in the battery charging process and the time-varying characteristics of the power grid environment. Through adaptive adjustment of neuron connection weights and activation thresholds, it achieves intelligent adaptive adjustment of the charging current. In summary, this invention solves the technical problem mentioned in the background art where charging piles cannot achieve intelligent adaptive adjustment of charging current in complex power grid environments. Attached Figure Description

[0018] Figure 1 This is a flowchart of the method of the present invention.

[0019] Figure 2 This is a schematic diagram of the liquid-state time-series regulation model.

[0020] Figure 3 The graph shows the change of the suggested output charging current value of the liquid timing regulation model in Example 2 over time.

[0021] Figure 4 This is a comparison chart of the charging current control curves for different types of batteries in Example 2.

[0022] Figure 5 This is a graph showing the performance index changes of the system in Example 2 over 6 months. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0024] like Figure 1 The diagram shown is a flowchart of a method for automatically adjusting the charging current of a high-power charging pile according to the first aspect of the present invention. This method includes the following steps:

[0025] S01. Construct a battery status monitoring matrix. By collecting parameters such as battery voltage, battery current, battery temperature, and battery internal resistance in real time, establish a basic data structure for battery status assessment. The number of rows in the battery status monitoring matrix corresponds to the types of monitoring parameters, and the number of columns corresponds to the time series.

[0026] S02. Establish a charging power offset matrix. Based on grid voltage fluctuations and load changes, calculate the deviation between the charging pile output power and the target power to form the data basis for power offset compensation.

[0027] S03. Construct a current regulation adjustable matrix. Based on the battery charging characteristic curve and safety threshold constraints, establish a dynamic adjustment range matrix for the charging current. The elements of the current regulation adjustable matrix represent the charging current adjustment boundary under different operating conditions.

[0028] S04. Generate a harmonic interference probability matrix, detect the harmonic content of the power grid through frequency domain analysis, calculate the interference probability of different frequency harmonics on the charging process, and establish a harmonic impact assessment model.

[0029] S05. Establish an upper-level game model with the goal of maximizing charging efficiency and a lower-level game model with the goal of optimizing battery safety. The two models achieve coordinated optimization through a charging current coupling function.

[0030] S06. The battery state monitoring matrix and the charging power offset matrix are processed using a pre-trained liquid timing adjustment model, and an optimized charging current adjustment suggestion value is output. The neuron connection weights of the liquid timing adjustment model are dynamically adjusted according to the battery temperature, voltage difference and charging time.

[0031] S07, optionally, further includes: calculating the final charging current output value through a charging current adaptive function, and adjusting the neuron activation threshold of the liquid timing adjustment model using different gain coefficients and damping parameters when the charging current adjustment recommendation value is in different ranges.

[0032] The battery state monitoring matrix stores and organizes time-series data of key battery state parameters. Each matrix element represents the measured values ​​of battery voltage, battery current, battery temperature, and battery internal resistance at a given time. Multi-parameter fusion analysis is achieved through matrix operations. The charging power offset matrix quantifies the difference between the actual output power and the ideal output power of the charging pile, providing data support for power adjustment. Each row of the charging power offset matrix corresponds to a different power level, and each column corresponds to a different compensation strategy. The adjustable current adjustment matrix defines the safe adjustment range of the charging current. The elements of the adjustable current adjustment matrix are determined based on battery type, charging stage, and environmental conditions to ensure that the charging process is carried out within safe boundaries. The harmonic interference probability matrix assesses the potential impact of grid harmonics on charging equipment. By statistically analyzing historical harmonic data and real-time monitoring results, the probability of occurrence and the degree of impact of harmonic interference are predicted.

[0033] The objective function of the upper-level game model is the charging efficiency maximization function. Its inputs include the charging pile output power, the suggested charging current adjustment value, the charging time, and the battery voltage. The output is the optimized charging efficiency value. Constraints include that the charging power does not exceed the device's rated power, the charging current is within a safe range, and the battery temperature is below a safe threshold. The objective function of the lower-level game model is the battery safety optimization function. Its inputs include the battery temperature, the suggested charging current adjustment value, the battery internal resistance, and the voltage difference. The output is the battery safety assessment value. Constraints include that the battery voltage does not exceed the cutoff voltage, the charging current change rate is limited, and the battery internal resistance change is within a normal range. The charging current coupling function coordinates the optimization processes of the upper-level and lower-level game models. Its inputs include the optimized charging efficiency value and the battery safety assessment value. The output is the coordinated charging current setting reference value.

[0034] The liquid-state time-series regulation model is structured as a recurrent neural network architecture based on a liquid state machine, comprising an input layer, a liquid reservoir, a readout layer, and a feedback connection layer. The liquid reservoir is composed of a large number of randomly connected neurons, whose connection weights and time constants are dynamically adjusted according to the frequency characteristics of the input signal. The liquid reservoir employs a leakage integrator neuron model, exhibiting memory decay and nonlinear activation characteristics. The readout layer extracts the liquid reservoir's state information through linear combination. The feedback connection layer feeds back the output information to the liquid reservoir to enhance time-series modeling capabilities. The training dataset establishment steps for the liquid-state time-series regulation model include collecting charging data from different types of batteries under various operating conditions, including battery voltage, battery current, battery temperature, battery internal resistance, and the corresponding optimal charging current setting. The data covers charging modes such as normal charging, fast charging, and trickle charging, and also includes charging data under different ambient temperature and humidity conditions. Preprocessing of the raw data includes denoising, normalization, and feature extraction. A correspondence between the input feature vector and the target output vector is established. The data is then divided into training, validation, and test sets according to the time series. The training steps of the liquid-state time-series regulation model include initializing the connection weights and time constants of the neurons in the liquid reservoir, training the dynamic characteristics of the liquid reservoir using an unsupervised learning method, stimulating the liquid reservoir to generate rich dynamic response patterns by inputting historical charging data sequences, training the linear weights of the readout layer using ridge regression, optimizing the model to minimize the mean square error between the predicted charging current and the actual optimal charging current, adjusting the model hyperparameters, including the liquid reservoir size, spectral radius, and input scaling factor, using cross-validation, and fine-tuning the weights of the feedback connection layer using the backpropagation algorithm to improve the time-series prediction accuracy.

[0035] The adaptive charging current function is used to calculate the final charging current output value based on the output of the liquid-state timing adjustment model and the current system state. The inputs include the suggested charging current adjustment value, the battery voltage, the grid quality index, and the ambient temperature. The output is the adaptively adjusted charging current output value. When the suggested charging current adjustment value is within the safe charging range, a small gain coefficient and a moderate damping parameter are used to adjust the neuron activation threshold to maintain a stable output. When the suggested charging current adjustment value is close to the safe boundary, a large gain coefficient and a strong damping parameter are used to increase the neuron activation threshold to enhance the system's conservatism. When the suggested charging current adjustment value exceeds the safe range, the maximum gain coefficient and the strongest damping parameter are used to significantly increase the neuron activation threshold to ensure system safety.

[0036] The specific implementation methods of the above steps are described in detail below.

[0037] The specific implementation of step S01 involves establishing a battery state monitoring matrix using a multi-sensor data acquisition system. This system employs a time-series data organization method to sample and store key battery parameters at fixed time intervals. First, a high-precision voltage sensor is used to collect battery voltage data in real time at 100-millisecond sampling intervals, achieving a sampling accuracy of 1 millivolt and covering a voltage measurement range of 0 to 1000 volts. Next, a Hall effect current sensor is used to collect battery charging current data, with the sampling frequency synchronized with the voltage. The current measurement accuracy is 1 milliampere, and the measurement range is 0 to 500 amperes. Simultaneously, a multi-point temperature sensor array is deployed to monitor temperature changes at different locations on the battery, with a temperature sampling accuracy of 0.1 degrees Celsius and a monitoring range of -20 degrees Celsius to 85 degrees Celsius. Battery internal resistance is measured using the AC impedance method, measuring changes in battery internal resistance by injecting a small-amplitude, high-frequency AC signal at a measurement frequency of 1 kHz and an internal resistance measurement accuracy of 0.1 milliohms. The monitoring matrix has four rows corresponding to voltage, current, temperature, and internal resistance parameters, respectively. The number of columns is determined based on the monitoring duration, generating 10 data points per second. The matrix uses a sliding window mechanism to dynamically update the data, with the window length set to 300 seconds, to ensure that the system can capture short-term fluctuations and long-term trend changes in battery status.

[0038] The specific implementation of step S02 is to establish a charging power offset matrix based on a power deviation analysis algorithm. This algorithm quantifies the power offset by using real-time power monitoring and target power comparison. The system first measures the actual output power of the charging pile in real time through a power metering module. The power measurement accuracy is 0.2% of the total power, and the measurement frequency is 20 times per second. The target power is dynamically set according to battery charging demand and grid load conditions. The calculation of the target power considers factors such as battery state of charge, charging mode selection, and grid power supply capacity. The power offset is calculated by the difference between the actual power and the target power. The number of rows in the offset matrix corresponds to different power levels, which are divided into low-power mode (0-30 kW), medium-power mode (30-120 kW), and high-power mode (120-350 kW). The number of columns corresponds to different compensation strategies, including voltage compensation, current limiting, frequency adjustment, and load balancing. The matrix element values ​​represent the effectiveness weights of different compensation strategies at each power level, with weights ranging from 0 to 1. The weight values ​​are determined through historical data statistical analysis and machine learning methods. The system uses exponential smoothing to filter the power offset data, with a smoothing coefficient set to 0.3, effectively reducing the impact of measurement noise on offset calculation.

[0039] The specific implementation of step S03 involves constructing an adjustable current regulation matrix using constrained optimization theory. This theory establishes the feasible region for current regulation based on battery charging characteristic curves and multiple safety constraints. The system first determines the basic charging characteristics of the battery using a battery type identification algorithm, supporting automatic identification of mainstream battery types such as lithium iron phosphate, ternary lithium, and lithium titanate. Corresponding mathematical models of charging characteristics are established for different battery types, with model parameters including maximum charging current, cutoff voltage, temperature coefficient, and aging factor. Safety threshold constraints include a battery temperature not exceeding 60 degrees Celsius, a single-cell voltage not exceeding 110% of the rated voltage, and a charging current change rate not exceeding 5% of the current current value per second. The number of rows in the adjustable matrix corresponds to different ranges of battery state of charge, divided into three stages: 0% to 20%, 20% to 80%, and 80% to 100%. The number of columns corresponds to different combinations of environmental conditions, including temperature range, humidity level, and ventilation conditions. The matrix elements represent the upper and lower limits of current regulation under corresponding operating conditions. The regulation range is corrected by a safety margin coefficient, which is dynamically adjusted based on battery aging and usage history, ranging from 0.8 to 1.0. The system employs fuzzy logic control to handle uncertainties in boundary conditions, ensuring a smooth transition of current regulation within the safety boundaries.

[0040] The specific implementation of step S04 involves generating a harmonic interference probability matrix using frequency domain analysis technology. This technology is based on the Fast Fourier Transform (FFT) algorithm for spectral analysis of grid voltage and current signals. The system is equipped with a high sampling rate data acquisition module, with a sampling frequency set to 10.24 kHz, capable of accurately capturing the 50 Hz fundamental wave and its harmonic components up to the 50th order. The harmonic detection algorithm employs a sliding window FFT method with a window length of 2048 sampling points and a window overlap rate of 50%, ensuring an optimal balance between time and frequency resolution in the spectral analysis. The system focuses on monitoring low-order harmonics (2nd, 3rd, 5th, 7th, 11th, and 13th harmonics and high-order harmonic groups. The harmonic content is calculated based on the total harmonic distortion (THD) index, which should be less than 5% during normal operation. The probability matrix is ​​established using statistical methods, establishing a predictive model for harmonic interference by analyzing the probability distribution characteristics of historical harmonic data. The number of rows in the matrix corresponds to different harmonic frequency components, and the number of columns corresponds to different interference intensity levels, categorized as slight interference (0-2%), moderate interference (2-5%), and severe interference (over 5%). Matrix element values ​​represent the probability of occurrence of each harmonic frequency at different intensity levels. Probability calculations employ Bayesian inference combined with real-time monitoring data and historical statistical data. The system also considers the interaction and superposition effects between harmonics, establishing a quantitative model of the impact of harmonic interference on the charging process through nonlinear regression analysis.

[0041] The specific implementation of step S05 involves establishing a two-layer optimization model using hierarchical game theory. This theory achieves multi-objective coordinated optimization of charging efficiency and battery safety through the Stackelberg game framework. The upper-layer game model aims to maximize charging efficiency, employing a gradient ascent algorithm to find the optimal solution. The input parameters of the objective function include the charging pile output power, the recommended charging current adjustment value, the charging time, and the battery voltage. The calculation of charging efficiency considers factors such as energy conversion loss, heat loss, and time cost, with an efficiency target value set above 92%. Constraints are handled using the Lagrange multiplier method to ensure that the charging power does not exceed 95% of the device's rated power, the charging current is within a safe range, and the battery temperature is below the safety threshold of 55 degrees Celsius. The lower-layer game model aims to optimize battery safety, using a particle swarm optimization algorithm to find the optimal strategy. The input parameters of the objective function include battery temperature, the recommended charging current adjustment value, battery internal resistance, and voltage difference. Safety assessment is based on a multivariate risk assessment model, comprehensively considering factors such as thermal runaway risk, overvoltage risk, and accelerated aging risk. The constraints include that the battery voltage does not exceed 98% of the cutoff voltage, the charging current change rate does not exceed 10% of the rated current per minute, and the battery internal resistance change is within 5% of the normal range. The charging current coupling function adopts a weighted comprehensive evaluation method to coordinate the optimization results of the two-layer game model. The weight coefficients are dynamically adjusted according to the charging stage and battery state. In the initial stage, safety is the main factor with a weight of 0.7, and in the later stage, efficiency is the main factor with a weight of 0.6.

[0042] The specific implementation of step S06 involves using a liquid-state time-series regulation model to process multidimensional state data and output optimized charging current regulation recommendations. This model, based on liquid state machine theory and reservoir calculation principles, achieves nonlinear mapping and prediction of time-series data. The model receives the battery state monitoring matrix and charging power offset matrix as inputs. The input layer performs standardized preprocessing on the multidimensional data, using zero-mean unit variance normalization. The liquid reservoir contains 500 leakage integrator neurons. The connections between neurons adopt a small-world network topology, with a connection probability set to 10%. The connection weights follow a normal distribution with a mean of 0 and a standard deviation of 0.5. The time constant of the neurons is adaptively adjusted according to the frequency characteristics of the input signal, ranging from 10 milliseconds to 100 milliseconds. The adjustment strategy is based on the degree of matching between the signal frequency and the neuron's characteristic frequency. The activation function of the reservoir uses a hyperbolic tangent function, and the activation threshold is dynamically adjusted based on battery temperature, voltage difference, and charging time. The readout layer uses linear regression to extract charging current regulation information from the reservoir state. The regression weights are obtained through ridge regression training, and the regularization parameter is set to 0.01. The feedback connection layer feeds the output information back to the reservoir input at a ratio of 0.1, enhancing the model's ability to model temporal dependencies. The model outputs a suggested charging current regulation value ranging from 0 to 500 amperes, with an output accuracy of 0.1 amperes.

[0043] The specific implementation of step S07 involves intelligently adjusting the final charging current output value through a charging current adaptive function. This function employs a piecewise linear control strategy to adjust the neuron activation parameters based on the range of the suggested value. The function inputs include the suggested charging current adjustment value output by the liquid-state timing regulation model, real-time battery voltage, grid quality indicators, and ambient temperature. The system first determines whether the suggested value is within the safe charging range, defined as 20% to 80% of the rated current. When the suggested value is within the safe range, the gain coefficient is set to 1.0, the damping parameter is set to 0.5, and the neuron activation threshold remains at the standard value of 0.5. When the suggested value approaches the safe boundary, i.e., within the range of 80% to 95% of the rated current, the gain coefficient is adjusted to 1.2, the damping parameter is increased to 0.7, and the activation threshold is raised to 0.7 to enhance system conservatism. When the suggested value exceeds the safe range, i.e., exceeds 95% of the rated current, the gain coefficient is set to the maximum value of 1.5, the damping parameter is set to the strongest value of 0.9, and the activation threshold is significantly raised to 0.9 to ensure system safety. The adaptive function also considers the impact of power grid quality indicators. When the power grid voltage fluctuation exceeds 5% of the rated value or the frequency deviation exceeds 0.5 Hz, all parameters are adjusted to a conservative 10%. The ambient temperature compensation mechanism adjusts the current output according to temperature changes; for every 10 degrees Celsius increase in temperature, the maximum output current decreases by 5%. The final output charging current value is smoothed and filtered to avoid abrupt changes, with a filtering time constant of 2 seconds.

[0044] The key technical ideas of this invention are mainly reflected in four aspects. First, the multi-dimensional state matrix fusion technology, by establishing a battery state monitoring matrix, a charging power offset matrix, a current regulation adjustable matrix, and a harmonic interference probability matrix, achieves unified modeling and collaborative analysis of multi-source heterogeneous data from the charging system. Compared to traditional control strategies that only consider a single or a few parameters, this technology can comprehensively capture various factors affecting the charging process and their interrelationships, significantly improving the accuracy and robustness of control decisions. Traditional control methods are often based on fixed control rules or simple feedback control, making it difficult to cope with complex and ever-changing charging environments. The multi-dimensional matrix fusion technology, through matrix operations, achieves multi-parameter coupled analysis, enabling the identification of nonlinear relationships and time-varying characteristics between parameters.

[0045] Secondly, the two-layer game theory optimization technique achieves a dynamic balance between efficiency and safety by constructing an upper-layer game theory model that maximizes charging efficiency and a lower-layer game theory model that optimizes battery safety. Compared to traditional single-objective optimization methods, the two-layer game theory model maximizes charging efficiency while ensuring battery safety, avoiding the problem of efficiency and safety being mutually restrictive in traditional methods. Traditional charging control strategies typically employ either conservative safety strategies or aggressive efficiency strategies, making it difficult to find the optimal balance between the two. The two-layer game theory technique establishes a hierarchical decision-making mechanism through the Stackelberg game theory framework. The upper-layer model pursues optimal efficiency, while the lower-layer model ensures a safety baseline. The two models achieve coordinated optimization through coupling functions, ensuring that the optimal control strategy is obtained under different charging stages and operating conditions.

[0046] Thirdly, there is the liquid-state time-series conditioning technology. The neural network model built upon liquid state machine theory possesses powerful time-series modeling and nonlinear mapping capabilities. Compared to traditional feedforward neural networks or simple recurrent neural networks, the liquid-state time-series model achieves efficient processing of time-series data and capture of long-term dependencies through reservoir computation mechanisms. Traditional neural networks suffer from gradient vanishing and training difficulties when processing time-series data during the charging process, while the random connection structure and dynamic weight adjustment mechanism of liquid reservoirs can naturally process time-series information, achieving good predictive performance without complex training processes. This technology is particularly suitable for processing non-stationary time-series data during the charging process and can adapt to long-term trends such as battery aging and environmental changes.

[0047] Fourth is the adaptive threshold adjustment technology. By dynamically adjusting the neuron activation threshold and control parameters according to different ranges of the recommended charging current, it achieves intelligent adjustment of the system response characteristics. Compared with the traditional fixed threshold control method, the adaptive threshold technology can dynamically adjust the conservatism of the control strategy according to changes in the system state, maximizing system performance while ensuring safety. Traditional control systems typically adopt a fixed conservative strategy to ensure safety, resulting in limited system performance under safe operating conditions. The adaptive threshold technology achieves adaptive optimization of the control strategy through piecewise linear control strategy and dynamic parameter adjustment. It adopts an aggressive strategy to improve efficiency within the safe range and a conservative strategy to ensure safety at dangerous boundaries.

[0048] The synergistic effect of these four key technological approaches forms a complete intelligent charging control system. Multi-dimensional matrix fusion provides the system with comprehensive and accurate state information; two-layer game theory optimization establishes a scientifically sound decision-making mechanism; liquid-phase timing regulation achieves high-precision control prediction; and adaptive threshold adjustment ensures the system's safe and reliable operation. Compared to existing technologies, this collaborative system can achieve a dynamic balance between charging efficiency and battery safety in complex and ever-changing charging environments, significantly improving the intelligence and adaptability of the charging system and providing a new technological path for the development of high-power charging technology.

[0049] It should be noted that the detailed structure of the liquid time-series regulation model is constructed based on reservoir calculation theory and the principle of liquid state machine, such as... Figure 2 As shown, the model comprises four core components. The input layer receives and preprocesses multi-dimensional time-series data, including the battery state monitoring matrix and charging power offset matrix. A sliding time window mechanism is used to organize the data temporally, with a window length of 50 time steps and a step size of 100 milliseconds. The liquid reservoir is the core computational unit of the model, consisting of 500 leakage integrator neurons. These neurons are sparsely and randomly connected with a connection density of 10%, and the connection weights are initialized with a Gaussian distribution. The neurons within the reservoir have different time constants, ranging from 10 milliseconds to 100 milliseconds, forming multi-scale temporal dynamics. The readout layer extracts reservoir state information through linear combination, with the weight matrix dimension being the output dimension multiplied by the number of reservoir neurons. The feedback connection layer feeds the output information back to the reservoir input at a ratio of 0.1, forming a closed-loop dynamic system to enhance the temporal modeling capability.

[0050] The establishment of the training dataset involves five detailed steps. The data collection phase requires acquiring charging data for different types of batteries under various operating conditions, covering mainstream battery types such as lithium iron phosphate, ternary lithium, and lithium titanate. Operating conditions include normal charging, fast charging, and trickle charging modes, while environmental conditions cover a temperature range of -20°C to 60°C and humidity range of 20% to 90%. The data preprocessing phase involves quality checks and cleaning of the raw data, removing outliers and missing values, using median filtering to remove high-frequency noise, and standardizing the data using the zero-mean, unit variance method. The feature extraction phase extracts statistical, frequency domain, and temporal features from the raw time-series data, including features such as mean, variance, peak value, and spectral energy distribution. The data partitioning phase divides the dataset into training, validation, and test sets according to the continuity of the time series, in a ratio of 6:2:2, ensuring temporal independence between different datasets. The label generation phase generates corresponding optimal charging current settings as training targets based on expert experience and historical best charging strategies.

[0051] The training process employs a phased training strategy. In the reservoir initialization phase, the spectral radius of the neuron connection weights is set to 0.95, and the input scaling factor is set to 0.1 to ensure the reservoir exhibits edge-stable dynamic characteristics. In the unsupervised pre-training phase, historical charging data is used to stimulate rich dynamic response patterns in the reservoir, and principal component analysis is used to evaluate the diversity and expressive power of the reservoir state. In the supervised learning phase, ridge regression is used to train the readout layer weights, with regularization parameters determined through cross-validation. The objective function is the mean square error between the predicted current and the actual optimal current. In the feedback weight optimization phase, gradient descent is used to fine-tune the feedback connection weights, with a learning rate set to 0.001 and 1000 training epochs. In the model validation phase, model performance is evaluated on an independent test set, with evaluation metrics including prediction accuracy, response time, and stability.

[0052] The liquid-state temporal regulation model is particularly suitable for solving the technical problems of this invention, mainly because the charging process has strong temporal dependence and nonlinear characteristics. The state changes during battery charging exhibit a significant memory effect; the current state depends not only on the immediate input but also on historical states. Traditional feedforward neural networks lack temporal memory capabilities and cannot effectively capture the dynamic characteristics of the charging process. While recurrent neural networks possess temporal modeling capabilities, they suffer from the vanishing gradient problem when processing long sequences, and the training process is complex. The liquid-state temporal model naturally achieves temporal memory through the random recursive connections of the reservoir, eliminating the need for complex gradient backpropagation training. Only a simple linear readout layer needs to be trained, thus avoiding the vanishing gradient problem.

[0053] Compared to the closest existing technologies, fuzzy control and traditional PID control, the liquid-state time-series regulation model has significant advantages. Fuzzy control relies on expert experience to formulate control rules, making it difficult to handle complex nonlinear relationships and time-varying characteristics, and its control accuracy is limited. PID control, based on linear control theory, cannot adapt to the strong nonlinear characteristics of the charging process, and its parameter tuning is difficult and its adaptability is poor. The liquid-state time-series model learns complex mapping relationships through a nonlinear dynamic system, automatically adapting to different battery types and operating conditions without manual parameter tuning. The reservoir computation mechanism gives the model strong generalization ability and robustness, enabling it to handle operating condition combinations not seen in the training data. The dynamic weight adjustment mechanism allows the model to automatically adjust the control strategy as the battery ages and the environment changes, achieving long-term stable control performance. The model has excellent real-time performance and low inference computation complexity, making it suitable for online real-time control applications.

[0054] It should be noted that this invention also solves the following technical problem: the lack of effective identification and compensation capabilities for grid harmonic interference in traditional charging pile systems. During actual charging, various frequency harmonics present in the power grid can interfere with charging equipment, affecting the stability of the charging current and charging efficiency. This invention constructs a harmonic interference probability matrix, uses frequency domain analysis to detect the grid harmonic content in real time, calculates the interference probability of different frequency harmonics on the charging process, and establishes a harmonic impact assessment model, providing a basis for harmonic compensation in charging control strategies. The liquid-state time-series adjustment model can learn and identify the time-varying characteristics of harmonic interference, and achieves intelligent compensation and suppression of harmonic interference by dynamically adjusting the neuron connection weights and activation thresholds.

[0055] Furthermore, this invention addresses the technical problem of existing charging control systems' inability to achieve coordinated optimization of charging efficiency and battery safety. Traditional control methods often employ a single optimization objective, either pursuing maximum charging efficiency while neglecting battery safety, or overly conservatively protecting the battery at the expense of charging efficiency, making it difficult to find the optimal balance between the two. This invention establishes a two-layer game theory model architecture, with the upper-layer model focusing on charging efficiency optimization and the lower-layer model focusing on battery safety assurance. The two models achieve information exchange and strategy coordination through a charging current coupling function, maximizing charging efficiency while ensuring battery safety, thus avoiding the one-sidedness and limitations of traditional methods.

[0056] A second aspect of the present invention provides a computer-readable storage medium storing program instructions, which, when executed in a computer, are used to perform the above-described method for automatically adjusting the charging current of a high-power charging pile.

[0057] A third aspect of the present invention provides an automatic adjustment system for charging current of a high-power charging pile, comprising the aforementioned computer-readable storage medium. The system is any one of a computer, a server, or a microcontroller. The computer-readable storage medium is disposed within the system, and the system is provided with a microprocessor that executes the program instructions stored in the computer-readable storage medium.

[0058] Specifically, the principle of this invention is as follows: The key to solving the technical problem of charging piles' inability to achieve intelligent adaptive adjustment of charging current in complex power grid environments lies in constructing an intelligent control architecture based on multi-dimensional data fusion and two-layer game optimization. First, through the collaborative construction of the battery state monitoring matrix, charging power offset matrix, adjustable current adjustment matrix, and harmonic interference probability matrix, this invention establishes a comprehensive perception system covering the internal state of the battery, the external environment of the power grid, safety constraints, and interference impact assessment, providing a complete data foundation for intelligent decision-making. Second, the upper-layer game model aims to maximize charging efficiency, while the lower-layer game model aims to optimize battery safety. The two models achieve coordinated optimization through a charging current coupling function. This two-layer game architecture ensures both the high efficiency of the charging process and the safety of system operation, avoiding the conflict between efficiency and safety that may occur with traditional single-objective optimization. The liquid-state time-series adjustment model serves as the core controller, employing a recurrent neural network architecture based on a liquid state machine. Its liquid reservoir consists of a large number of randomly connected neurons, possessing rich dynamic response modes and powerful nonlinear mapping capabilities, enabling accurate modeling of the complex time-varying characteristics of the battery charging process. In the model, the neuron connection weights are dynamically adjusted based on battery temperature, voltage difference, and charging time. The activation threshold is optimized in real time according to the current system state through an adaptive charging current function, achieving adaptive adjustment of control parameters. This design enables the charging pile to automatically adjust its charging strategy based on real-time monitoring of battery status and grid environment, maintaining optimal charging performance even under complex and changing operating conditions.

[0059] The following provides a specific embodiment 1 of the present invention, and the specific implementation of each step in this embodiment 1 is described in detail below.

[0060] The specific implementation of step S01 involves constructing a battery state monitoring matrix, which is used to store time-series data of key battery parameters, as shown below:

[0061] ;

[0062] In the formula, Battery status monitoring matrix; For the first The battery voltage at any given moment, in volts; For the first The battery current at any given moment, measured in amperes; For the first Battery temperature at any given moment, in degrees Celsius; For the first The internal resistance of the battery at a given moment, in milliohms; For the first Each sampling time, ; This represents the total number of sampling points. The parameter acquisition method is as follows: Real-time data acquisition is achieved through a high-precision voltage sensor with a sampling frequency of 10 Hz and a measurement accuracy of 1 millivolt. The data is acquired using a Hall current sensor, with the sampling frequency synchronized with the voltage, and the measurement accuracy is 1 mA. Temperature is acquired through a multi-point temperature sensor array and measured using thermocouples, with a measurement accuracy of 0.1 degrees Celsius. The measurement was performed using the AC impedance method, and the result was calculated by injecting a small-amplitude AC signal with a frequency of 1 kHz.

[0063] The specific implementation of step S02 is to establish a charging power offset matrix. First, the deviation between the actual output power of the charging pile and the target power is calculated, as follows:

[0064] ;

[0065] In the formula, For a moment The power deviation value, in kilowatts; This refers to the actual output power of the charging pile, in kilowatts. The target output power is expressed in kilowatts. The noise term is measured in kilowatts. The charging power offset matrix is ​​represented as follows:

[0066] ;

[0067] In the formula, This is the charging power offset matrix; For the first The power level The weighting coefficients of the compensation strategies These correspond to low, medium, and high power modes, respectively. These correspond to voltage compensation, current limiting, frequency regulation, and load balancing strategies, respectively. The parameter acquisition method is as follows: The power is measured in real time using a three-phase power measurement module, with a measurement accuracy of 0.2% of the total power. Determined based on battery state of charge and charging requirements; The standard deviation ranges from 0.1% to 0.5% of the measured value; Determined through historical data statistical analysis and least squares fitting.

[0068] The specific implementation of step S03 involves constructing an adjustable current regulation matrix, establishing a feasible region for current regulation based on constrained optimization theory, and calculating the current regulation range using the following formula:

[0069] ;

[0070] In the formula, In the state of charge and environmental conditions The adjustable current value is expressed in amperes. The reference charging current is expressed in amperes. For safety factors, the value ranges from 0.8 to 1.0; This is the aging correction factor, with a value ranging from 0.85 to 1.0; This is a temperature compensation term, measured in amperes. The ambient temperature is expressed in degrees Celsius. The adjustable current matrix is ​​represented as follows:

[0071] ;

[0072] In the formula, It is an adjustable matrix for current regulation; These correspond to three ranges: 0% to 20%, 20% to 80%, and 80% to 100% of the state of charge. These correspond to different combinations of environmental conditions. The method for obtaining the parameters is as follows: Determined based on the battery charging characteristic curve; Calculated based on safety threshold constraints; Obtained through battery capacity degradation testing; The calculation was performed using linear interpolation of the temperature coefficient.

[0073] The specific implementation of step S04 involves generating a harmonic interference probability matrix, performing frequency domain analysis on the power grid signal using a fast Fourier transform, and calculating the total harmonic distortion using the following formula:

[0074] ;

[0075] In the formula, Total harmonic distortion (THD) is expressed as a percentage. For the first The amplitude of the subharmonic, measured in volts or amperes; This is the fundamental frequency amplitude, measured in volts or amperes. The highest harmonic order for analysis is typically taken as 50; The measurement error term ranges from 0.1% to 0.3%. The harmonic interference probability matrix is ​​expressed as:

[0076] ;

[0077] In the formula, This is the harmonic interference probability matrix; For the first Subharmonics in interference intensity level The probability of occurrence of the following, For harmonic order index, These correspond to mild, moderate, and severe interference levels, respectively. The parameter acquisition method is as follows: The calculation was performed using Fast Fourier Transform, with a sampling frequency of 10.24 kHz and a window length of 2048 points. Based on statistical analysis of historical harmonic data, Bayesian inference methods were used for calculation.

[0078] The specific implementation of step S05 is to establish a two-layer game optimization model, where the objective function of the upper-layer game model is:

[0079] ;

[0080] In the formula, The objective function is the charging efficiency. This refers to the output power of the charging station, measured in kilowatts (kW). These are recommended values ​​for charging current adjustment, in amperes. Charging time is expressed in hours. Battery voltage, measured in volts, originating from... Real-time value; Total input energy, expressed in kilowatt-hours; Heat loss, expressed in kilowatt-hours; For time cost, the equivalent energy unit kilowatt-hour is used; This is the penalty factor, with a value ranging from 0.01 to 0.1; For the first One constraint violation quantity; To constrain the total number, the objective function of the lower-level game model is:

[0081] ;

[0082] In the formula, The objective function is battery safety. Battery temperature, in degrees Celsius, sourced from The real-time value; The internal resistance of the battery is measured in milliohms, and is derived from... The real-time value; This is the voltage difference, measured in volts. This is a thermal safety assessment value; This is the voltage safety assessment value; This is the aging safety assessment value; These are the weighting coefficients; This is a risk penalty factor, with a value ranging from 0.05 to 0.2; For the first Each security risk quantity For risk indexing; This represents the total number of risk factors. The charging current coupling function is:

[0083] ;

[0084] In the formula, Set a reference value for the coordinated charging current, in amperes; These are the linear coupling coefficients; is the nonlinear coupling coefficient.

[0085] The specific implementation of step S06 is the same as that of the aforementioned liquid time-series regulation model training, and will not be described in detail here.

[0086] The specific implementation of step S07 is to calculate the final output value through a charging current adaptive function. This function adopts a piecewise linear control strategy, as shown below:

[0087] ;

[0088] ;

[0089] In the formula, This is the final charging current output value, in amperes. This is the rated charging current, measured in amperes. The gain coefficients for different ranges are 1.0, 1.2, and 1.5, respectively. This is the bias term, and the unit is amperes. For the first Environmental compensation function for each interval, For range index; This is the mains voltage, measured in volts. Here, represents the ambient temperature, expressed in degrees Celsius. The environmental compensation function is expressed as:

[0090] ;

[0091] In the formula, For the first The compensation coefficient for each interval; This is the rated voltage of the power grid, measured in volts. For reference temperature, the value is 25 degrees Celsius.

[0092] It needs to be explained that the battery state monitoring matrix Multidimensional time-series data is organized in matrix form, with row vectors representing different parameter types and column vectors representing time series, thus establishing the mathematical foundation for multi-parameter fusion analysis. Compared to traditional single-parameter monitoring methods, this matrix structure can capture the temporal correlation and coupling relationships between parameters, providing comprehensive state information for subsequent intelligent decision-making and significantly improving the accuracy and robustness of charging control.

[0093] Power deviation calculation formula By introducing a noise term Uncertainties in actual measurements are taken into account, making the power offset assessment more accurate and reliable. Charging power offset matrix. A mapping relationship between power level and compensation strategy was established. Compared with the traditional fixed compensation method, it can adaptively select the optimal compensation strategy according to different power levels, which effectively improves the accuracy and response speed of power regulation.

[0094] Formula for calculating current adjustment range This formula comprehensively considers multiple influencing factors, including reference current, safety factor, aging correction, and temperature compensation, and achieves the synergistic effect of multiple factors through a combination of product and addition. Compared with traditional single-factor adjustment methods, this formula can more comprehensively reflect the actual charging capacity and safety boundary of the battery, ensuring a dynamic balance between optimal efficiency and safety during the charging process.

[0095] Total Harmonic Distortion Calculation Formula Harmonic energy is calculated using the mathematical form of the sum of squares and the square root, and a measurement error term is introduced. Improved calculation accuracy. Harmonic interference probability matrix. A probabilistic mapping between harmonic frequency and interference intensity was established. Compared with traditional deterministic harmonic analysis methods, it can quantify the uncertainty of harmonic interference and provide a probabilistic decision-making basis for the formulation of charging control strategies.

[0096] The objective function of the two-level game model softens the constraints by introducing a penalty term; the upper-level efficiency objective function... The efficiency index is defined by a ratio, and the lower-level security objective function is... Multiple security factors are synthesized using a weighted summation method. Coupling function. Through linear terms and nonlinear terms The combination of these two objectives achieves coordinated optimization, and compared with traditional single-objective optimization methods, it can achieve Pareto optimality between efficiency and safety, significantly improving the overall performance of the charging strategy.

[0097] Adaptive charging current function A piecewise linear structure is adopted, and different gain coefficients and compensation strategies are used in different current ranges. Environmental compensation function. The influence of grid voltage and ambient temperature is considered through a linear combination. Compared with traditional fixed-parameter control methods, this adaptive mechanism can dynamically adjust the conservatism of the control strategy according to the system state, maximizing system performance while ensuring safety, thus realizing the intelligence and adaptability of the control strategy.

[0098] To better understand and implement this invention, a specific application scenario is provided below as Example 2: A city's electric vehicle charging station is equipped with 20 350 kW high-power charging piles, serving more than 500 electric vehicles daily. It faces technical challenges such as severe load fluctuations in the city's power grid during summer, diverse vehicle battery types, and complex environmental conditions. The technical team adopted the automatic charging current adjustment method of this invention to construct a complete intelligent charging control system.

[0099] The technical team first established a battery state monitoring matrix and deployed a high-precision sensor network to monitor the charging process in real time. The system collects four key parameters simultaneously—battery voltage, current, temperature, and internal resistance—at a sampling interval of 100 milliseconds. During a typical charging process, the voltage range of a lithium iron phosphate battery is 280 volts to 390 volts, the charging current ranges from 50 amps to 300 amps, the battery temperature is controlled between 15 degrees Celsius and 45 degrees Celsius, and the internal resistance varies from 0.8 milliohms to 1.5 milliohms. Ternary lithium batteries have different parameter ranges: a voltage range of 320 volts to 420 volts, a charging current reaching 350 amps, and a relatively narrower temperature range of 20 degrees Celsius to 40 degrees Celsius. The monitoring matrix established by the system contains 3000 data points every 300 seconds, providing rich foundational data for subsequent intelligent decision-making.

[0100] The charging power offset matrix was established based on in-depth analysis of the power grid conditions. The technical team found that the voltage fluctuations in the regional power grid were significant during morning and evening peak hours, with peak voltage reaching 108% of the rated value and valley voltage dropping to 92%. The power offset calculation results showed that the offset was typically between 2% and 5% in low-power mode, increased to 5% to 12% in medium-power mode, and exceeded 15% in high-power mode. The system formulated corresponding compensation strategy weights for different offset scenarios, as shown in Table 1.

[0101] Table 1 Weighting Table for Power Offset Compensation Strategy

[0102]

[0103] The adjustable current regulation matrix is ​​designed to take into account the specific requirements of different battery types and charging stages. Based on the battery charging characteristic curve, the technical team divided the charging process into three stages: the initial fast charging stage, the mid-term constant current stage, and the late trickle charging stage. The safety factor for the initial fast charging stage is set at 0.95, the aging correction factor at 0.90, and the temperature compensation range at -10 amps to +15 amps. In the mid-term constant current stage, the safety factor is increased to 0.98, the aging correction factor is 0.95, and the temperature compensation range is narrowed to -5 amps to +8 amps. The late trickle charging stage uses the most conservative parameter settings: a safety factor of 1.0, an aging correction factor of 1.0, and a temperature compensation range at -2 amps to +3 amps.

[0104] The generation of the harmonic interference probability matrix is ​​based on long-term monitoring of the harmonic characteristics of the power grid in this region. The technical team used a data acquisition system with a sampling frequency of 10.24 kHz to accurately analyze the harmonic components in the power grid. Monitoring results show that the total harmonic distortion rate of the power grid in this region is 3.2% under normal conditions, reaching 4.8% during peak load periods. The main harmonic components include the 3rd, 5th, 7th, and 11th harmonics, with the 5th harmonic being the most abundant, reaching 2.1% of the fundamental frequency. The harmonic interference probability matrix established by the system shows that the probability of minor interference is 65%, the probability of moderate interference is 28%, and the probability of severe interference is 7%.

[0105] The establishment of a two-layer game theory optimization model is the core technical aspect of the system. The upper-layer game theory model aims to maximize charging efficiency. Analysis shows that the energy conversion efficiency during a typical charging process can reach 93.5%, with heat loss accounting for 4.2% of the total energy and time cost equivalent energy loss at 2.3%. The system's constraints include charging power not exceeding 95% of the device's rated power, charging current not exceeding 98% of the battery's safety threshold, and battery temperature not exceeding 50 degrees Celsius. The lower-layer game theory model aims for optimal battery safety and establishes a comprehensive safety assessment system. The thermal safety assessment value considers battery temperature distribution, heat dissipation efficiency, and thermal runaway risk, with a typical range of 0.85 to 0.95. The voltage safety assessment value is based on single-cell voltage balance and overvoltage risk, with a typical range of 0.90 to 0.98. The aging safety assessment value comprehensively considers capacity decay rate and internal resistance growth trend, with a typical range of 0.88 to 0.96.

[0106] The liquid-phase time-series regulation model was trained using a large-scale dataset containing 50,000 charging cycles. The dataset covers charging data from 15 different types of electric vehicle batteries under various seasons, time periods, and load conditions. The model's liquid reservoir contains 500 neurons, employing a small-world network topology with a connection density of 10%. After training, the model achieved a prediction accuracy of 97.3% on the test set, with a response time of 15 milliseconds, and can output optimized charging current regulation suggestions in real time. Figure 3 As shown, the recommended charging current output by the model exhibits a clear phased characteristic as time changes. The recommended value is higher in the early stage of charging and gradually decreases as the state of charge increases, reflecting an accurate grasp of the battery charging characteristics.

[0107] The adaptive charging current function employs corresponding control strategies based on the range of the recommended value. When the recommended value is within the safe charging range, the system uses a standard control mode with a gain coefficient of 1.0 and a damping parameter of 0.5. When the recommended value approaches the safety boundary, the system automatically switches to a conservative control mode, adjusting the gain coefficient to 1.2 and increasing the damping parameter to 0.7. When the recommended value exceeds the safe range, the system activates the highest safety level control, setting the gain coefficient to 1.5 and increasing the damping parameter to 0.9. The environmental compensation mechanism dynamically adjusts based on grid voltage fluctuations and ambient temperature changes. For every 1% fluctuation in grid voltage, the current output adjusts by 2%, and for every 5 degrees Celsius change in ambient temperature, the maximum output current adjusts by 3%.

[0108] The actual performance after the system was put into operation verified the effectiveness of the technical solution. During a six-month operational monitoring period, the electric vehicle charging process served by the system demonstrated good stability and safety. The peak battery temperature during charging was reduced by 8% compared to traditional control methods, charging time was shortened by 12%, and energy utilization efficiency was improved by 5%. Figure 4 As shown, the system exhibits good adaptability during the charging process of different types of batteries. The charging curve of lithium iron phosphate batteries is smooth and stable, and the charging process of ternary lithium batteries also achieves a good balance between efficiency and safety. The interference of grid harmonics on the charging process is effectively suppressed, and the fluctuation of charging current is controlled within 3%.

[0109] The technical team also tested the system's performance under extreme conditions. In hot summer weather, when the ambient temperature reached 42 degrees Celsius, the system automatically adjusted its charging strategy, reducing the charging current and increasing the heat dissipation interval to ensure the battery temperature remained within a safe range. In cold winter weather, when the ambient temperature dropped to -15 degrees Celsius, the system ensured normal charging by increasing the initial charging current and extending the preheating time. Under sudden changes in grid load, the system's response time was within 50 milliseconds, enabling it to quickly adjust charging parameters to adapt to changes in grid conditions.

[0110] The system's maintenance and upgrades also reflect intelligent features. The liquid timing regulation model has online learning capabilities, continuously optimizing the control strategy based on new charging data. The technical team has set up an automatic data update mechanism; after collecting 1000 new charging cycles of data, the system automatically fine-tunes the model parameters. The weight coefficients of the two-layer game model are also periodically adjusted based on long-term operational statistics to ensure that the balance between efficiency and safety is always optimal. Figure 5 As shown, the system's performance indicators showed a continuous improvement trend after 6 months of operation, and all key parameters exhibited good convergence characteristics.

[0111] The statistical results of the multi-dimensional matrix fusion processing are shown in Table 2, demonstrating the system's comprehensive perception capability of complex charging environments.

[0112] Table 2 Statistical Table of Multidimensional State Parameters

[0113]

[0114] This invention represents a significant technological advancement over traditional charging control methods. Traditional charging control methods typically employ fixed control parameters and simple feedback control strategies, failing to adapt to complex and ever-changing charging environments and the personalized needs of different battery types. This invention, through multi-dimensional state matrix fusion technology, achieves comprehensive and multi-faceted real-time monitoring and analysis of the charging process. Compared to traditional single-parameter control methods, it captures richer state information and the coupling relationships between parameters. The application of a two-layer game optimization model overcomes the limitations of traditional single-objective optimization. By establishing a hierarchical decision-making mechanism, it maximizes charging efficiency while ensuring battery safety, avoiding the problem of efficiency and safety mutually constraining each other in traditional methods. The liquid-state time-series adjustment model, based on the time-series modeling capabilities of neural networks, better handles the nonlinear dynamic characteristics and long-term dependencies in the charging process compared to traditional PID control or fuzzy control methods, achieving accurate prediction and control of the charging current. The adaptive threshold adjustment technology dynamically adjusts the control strategy according to the system state. Compared to traditional fixed-threshold control methods, it can automatically optimize control parameters under different operating conditions, achieving intelligent and adaptive control strategies. The synergistic effect of these technological innovations has enabled the entire charging control system to exhibit greater robustness and adaptability in complex environments, providing a new technological path for the development of high-power charging technology.

[0115] ====↓↓↓↓↓====Variable Explanation====↓↓↓↓↓====

[0116] It should be noted that the variables involved in this invention are explained in detail in Tables 3 and 4 below.

[0117] Table 3. Variable Explanation Table (Part 1)

[0118]

[0119] Table 4. Variable Explanation Table (Part Two)

[0120]

[0121] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for automatically adjusting the charging current of a high-power charging pile, characterized in that, This includes constructing a battery status monitoring matrix, which establishes a basic data structure for battery status assessment by collecting parameters such as battery voltage, battery current, battery temperature, and battery internal resistance in real time; and constructing a charging power offset matrix, which calculates the deviation between the output power of the charging pile and the target power based on grid voltage fluctuations and load changes, forming a data basis for power offset compensation. An adjustable current regulation matrix is ​​constructed, and a dynamic adjustment range matrix for the charging current is established based on the battery charging characteristic curve and safety threshold constraints. A harmonic interference probability matrix is ​​generated, and the harmonic content of the power grid is detected by frequency domain analysis. The interference probability of different frequency harmonics on the charging process is calculated, and a harmonic impact assessment model is established. An upper-level game model with the goal of maximizing charging efficiency and a lower-level game model with the goal of optimizing battery safety are established. The two models are coordinated and optimized through a charging current coupling function. A pre-trained liquid-state timing regulation model is used to process the battery state monitoring matrix and the charging power offset matrix, and an optimized charging current regulation suggestion value is output. The final charging current output value is calculated through a charging current adaptive function. Among them, the objective function of the upper-level game model is the charging efficiency maximization function. The inputs include the charging pile output power, the suggested value of charging current adjustment, the charging time and the battery voltage. The output is the optimized value of charging efficiency. The constraints include the charging power not exceeding the rated power of the equipment, the charging current being within the safe range, and the battery temperature being below the safe threshold. The objective function of the lower-level game model is the optimal function for battery safety. The inputs include battery temperature, recommended value for charging current adjustment, battery internal resistance and voltage difference. The output is the battery safety assessment value. The constraints include battery voltage not exceeding the cutoff voltage, limited rate of change of charging current, and battery internal resistance change within the normal range. The structure of the liquid-state time-series regulation model is a recurrent neural network architecture based on a liquid state machine. The time constant of the neuron is adaptively adjusted according to the frequency characteristics of the input signal. The adjustment strategy is based on the degree of matching between the signal frequency and the neuron's characteristic frequency. The activation function of the liquid reservoir adopts the hyperbolic tangent function, and the activation threshold is dynamically adjusted according to the battery temperature, voltage difference, and charging time.

2. The automatic adjustment method for charging current of a high-power charging pile according to claim 1, characterized in that, The number of rows in the battery status monitoring matrix corresponds to the types of monitoring parameters, and the number of columns corresponds to the time series. It is used to store and organize time-series data of key battery status parameters. Each matrix element represents the measured values ​​of battery voltage, battery current, battery temperature, and battery internal resistance at a given time. Multi-parameter fusion analysis is achieved through matrix operations.

3. The automatic adjustment method for charging current of a high-power charging pile according to claim 2, characterized in that, The charging power offset matrix is ​​used to quantify the difference between the actual output power and the ideal output power of the charging pile, providing data support for power adjustment. Each row of the charging power offset matrix corresponds to a different power level, and each column corresponds to a different compensation strategy.

4. The automatic adjustment method for charging current of a high-power charging pile according to claim 3, characterized in that, The adjustable current matrix is ​​used to define the safe adjustment range of the charging current. The elements of the adjustable current matrix represent the charging current adjustment boundary under different operating conditions. The elements of the adjustable current matrix are determined according to the battery type, charging stage and environmental conditions to ensure that the charging process is carried out within the safe boundary.

5. The automatic adjustment method for charging current of a high-power charging pile according to claim 4, characterized in that, The harmonic interference probability matrix is ​​used to assess the potential impact of power grid harmonics on charging equipment. By statistically analyzing historical harmonic data and real-time monitoring results, it predicts the probability of occurrence and the degree of impact of harmonic interference.

6. The automatic adjustment method for charging current of a high-power charging pile according to claim 5, characterized in that, The charging current coupling function is used to coordinate the optimization process of the upper-level game model and the lower-level game model. The inputs include the optimized charging efficiency value and the battery safety assessment value, and the output is the coordinated charging current setting reference value.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores program instructions, which, when executed in a computer, are used to perform the automatic adjustment method for charging current of a high-power charging pile as described in any one of claims 1-6.

8. A high-power charging pile charging current automatic adjustment system, characterized in that, The system includes the computer-readable storage medium of claim 7, wherein the system is any one of a computer, a server, or a microcontroller, the computer-readable storage medium is disposed within the system, and the system is provided with a microprocessor that executes the program instructions stored in the computer-readable storage medium.

Images