Control method of charging pile equipment

Abstract

The invention relates to the technical field of electric vehicle charging pile regulation and control, in particular to a control method of charging pile equipment, which comprises the following steps of: constructing a multi-mode sensing data fusion model, integrating electrochemical impedance spectroscopy, ultrasonic wave and infrared thermal imaging sensor data, deeply excavating the aging condition of a battery, and obtaining the charging pile equipment. The aging trend is predicted by means of a deep belief network, a dynamic optimal power strategy is generated according to a deep Q network algorithm, the charging efficiency is improved, the service life of the battery is prolonged, and for single battery management, an electrochemical-thermal-mechanical coupling microenvironment model and a finite element and data driving fusion strategy are used for solving the problem. The microenvironment of the single batteries is strictly monitored and finely balanced and regulated, and the temperature and voltage differences among the single batteries are strictly controlled in an extremely small range, so that the performance consistency and stability of the battery pack are greatly improved, and the charging efficiency of the single batteries of the electric vehicle is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle charging pile control, and in particular to a control method for a charging pile device. Background Art

[0002] Electric vehicle charging pile regulation is an important technology. The electric vehicle industry is booming. Charging piles are key supporting facilities, and their control technology has a profound impact on battery performance and life. Traditional charging pile control methods have significant shortcomings when dealing with battery aging problems.

[0003] When adjusting the power, traditional charging piles mainly rely on basic parameters such as voltage and current on the battery surface, and lack deep insight into the complex electrochemical mechanisms behind battery aging. Battery aging causes changes in the structure of electrode materials and degradation of electrolyte performance. Traditional methods cannot accurately quantify the demand for charging power caused by these changes. This rough power regulation makes the charging power of electric vehicle batteries inaccurate. If it is too high, it is easy to cause irreversible damage such as lithium deposition and overheating inside the battery, accelerating aging. If it is too low, it will greatly extend the charging time and reduce the use efficiency. In order to solve this technical problem, we provide a control method for charging pile equipment. Summary of the invention

[0004] The purpose of the present invention is to provide a control method for a charging pile device to solve the problems raised in the above background technology.

[0005] To achieve the above object, one of the objects of the present invention is to provide a control method for a charging pile device, comprising the following steps: S1. Deploy electrochemical impedance spectroscopy sensors, ultrasonic sensor arrays, and infrared thermal imaging sensor networks. Electrochemical impedance spectroscopy sensors are used to obtain data reflecting the internal electrochemical state of the battery. Ultrasonic sensor arrays are used to detect battery plate deformation, stratification, and electrolyte distribution. Infrared thermal imaging sensors are used to obtain battery surface temperature distribution. S2. Build a multimodal data fusion model based on a deep learning framework, synchronize the data collected by the electrochemical impedance spectroscopy sensor, the data collected by the ultrasonic sensor array, and the infrared thermal imaging sensor monitoring data in time and space, and fuse them at the feature level, and output a feature vector that comprehensively reflects the battery aging status; S3. Collect multimodal sensing data of electric vehicle batteries under different charge and discharge rates, ambient temperature and humidity conditions and the corresponding battery remaining life and capacity decay curve data, train a battery aging prediction model based on a deep belief network, predict the aging evolution trend of the battery during the current charging process based on real-time multimodal sensing data, and generate the optimal power adjustment strategy using a deep Q network algorithm based on the aging evolution trend output by the battery aging prediction model, combined with the current state of charge, temperature distribution and real-time power capacity of the charging pile; S4. Establish an electrochemical-thermal-mechanical coupling microenvironment model for each single cell in the battery pack. By arranging a network of micro temperature, pressure, and voltage sensors in the battery pack, collect the changes in the microenvironment parameters of each single cell. Use a combination of finite element analysis and data-driven methods to model and simulate the internal current distribution, temperature field distribution, and stress distribution of the single cell in real time, and adjust the charging current path and power distribution of the charging pile to each single cell.

[0006] As a further improvement of the technical solution, the method for spatiotemporal synchronization of multimodal data in S2 is as follows: The sensor network is equipped with a clock synchronization module based on GPS timing. Each sensor has a built-in GPS receiving chip to receive satellite signals to obtain a timestamp, which is used as the starting time reference for data collection. When all sensors receive a specific synchronization pulse in the GPS signal, the data acquisition circuit is triggered synchronously to record electrochemical impedance spectroscopy data, ultrasonic data, and infrared thermal imaging data from that moment on. After completing data collection, a time deviation correction model is established to analyze the transmission delay of data from the sensor to the fusion processing center. By embedding a time monitoring module in the key nodes of the transmission link, the sending and receiving timestamps of the data packets are collected, the transmission delay is calculated, and the least squares method is used to fit the relationship model between the transmission delay and the data flow and link load, and the transmission delay of each data packet is predicted and compensated. At the same time, the standard clock source signal is regularly collected to calibrate the local clock deviation of the sensor, and the timestamp of the collected data is corrected by linear interpolation.

[0007] As a further improvement of the technical solution, the method for feature-level fusion of multimodal data in S2 is as follows: The electrochemical impedance spectroscopy data, ultrasonic data and infrared thermal imaging data were normalized to obtain the energy characteristics of each frequency band, marginal spectrum characteristics, and the temperature mean, variance and temperature gradient characteristics of each region; A deep convolutional neural network fusion architecture is constructed, which includes an input layer, a convolutional layer, a pooling layer and a fully connected layer. The input layer receives a normalized multimodal feature vector, the convolutional layer is used to extract features, the pooling layer is used to reduce the dimension of the extracted features, and the fully connected layer is used to output a feature vector that comprehensively reflects the probability distribution of battery aging.

[0008] As a further improvement of the technical solution, the sample preprocessing and feature optimization method for deep belief network training in S3 is as follows: Normalize the eigenvector that comprehensively reflects the probability distribution of battery aging degree with the collected multimodal sensing data and battery life-related data, calculate the data covariance matrix and solve the eigenvalues ​​and eigenvectors, select the main components according to the eigenvalues, and project the high-dimensional data into a low-dimensional space; The mutual information value between each feature and the battery remaining life and capacity decay curve data is calculated, and the key features are selected according to the mutual information value. The mutual information value is used to measure the correlation between the feature and the target variable.

[0009] As a further improvement of the technical solution, the environment modeling and strategy optimization method of the deep Q network algorithm in S3 is as follows: The battery charging process is abstracted into a Markov decision process environment model, and the state space, action space and transition probability are defined. The state space includes the battery state of charge, temperature distribution and aging degree. The action space is the charging pile power adjustment strategy. The transition probability is used to describe the probability of transferring to another state after performing an action in a certain state. Then, the reward function is obtained based on historical data and model calculation. The deep Q network is divided into a main network and a target network. The main network is used to select action evaluation values, and the target network is used to calculate target Q values. Priorities are then set for historical multimodal perception data. The priorities are calculated by a preset fixed formula. Sample data whose priorities exceed the preset priority threshold are replayed for training to complete the optimization.

[0010] As a further improvement of the technical solution, the multi-model fusion prediction method of the battery aging evolution trend in S3 is as follows: Fusion of long short-term memory network and deep belief network, wherein the long short-term memory network is used to process multimodal perception data time series to capture dynamic features, and the deep belief network is used to learn data abstract features and aging relationships; During fusion, the output sequence features of the long short-term memory network and some hidden layer outputs of the deep belief network are spliced ​​into a joint feature vector and input into the subsequent layer. Then, based on the deep belief network's prediction of the aging trend, the support vector machine regression model is introduced to learn the prediction residual correction. The deep belief network is trained with one-third of the data to obtain the preliminary prediction value, and the residual is calculated. Finally, the support vector machine regression model is trained with the residual to obtain the correction function, and the final prediction value is obtained based on this.

[0011] As a further improvement of the technical solution, the parameter adaptive identification method of the electrochemical-thermal-mechanical coupling microenvironment model in S4 is as follows: The extended Kalman filter algorithm is used to identify the key parameters of the model in real time, and the state equation and measurement equation are constructed, and the state estimation and covariance are initialized. The state equation is used to predict the state of the system at the next moment, and the measurement equation is used to describe the relationship between the sensor and the system state. At each sampling moment, the prediction step calculates the prior estimate and covariance, and the update step calculates the Kalman gain based on the measurement value and updates the state estimate and covariance.

[0012] As a further improvement of the technical solution, the modeling and simulation control method of the finite element analysis and data-driven fusion in S4 is as follows: Constructing a three-dimensional solid model according to the battery geometry and material properties, dividing the grid, and setting control equations, wherein the control equations include heat conduction equations, electrical conduction equations, and mechanical equilibrium equations; Redefine the boundary conditions, collect the battery microenvironment parameters and charge and discharge performance data under multiple working conditions, the microenvironment parameters include temperature, pressure and voltage, build an artificial neural network, the input parameters are sensor data, the output is the temperature prediction value of the temperature field node, the prediction value is used to calibrate the finite element model parameters; During the charging process, the model input is updated based on the real-time collected sensor data, the finite element equations and data-driven model predictions are solved in real time, the dynamic current, temperature, and stress distributions are obtained, a multi-objective optimization function is constructed, and a genetic algorithm is used to calculate the optimal charging current path and power allocation strategy, so that the multi-objective optimization function is minimized while satisfying the battery safety constraints.

[0013] Compared with the prior art, the present invention has the following beneficial effects: In a control method for charging pile equipment, in terms of power regulation, by deploying multiple types of sensors to build a multi-modal perception data fusion system, the battery aging status can be keenly captured, the deep neural network model accurately calculates the aging trend, and the deep Q network algorithm generates a dynamic power strategy, which greatly improves the adaptability of charging power. Based on the coordinated and balanced control of the microenvironment of the single cell battery, the temperature and voltage differences between the single cells are successfully and strictly controlled within a very small range by coupling the microenvironment model and intelligent control means, ensuring that the charging status of each single cell is uniform and consistent, the consistency of the battery pack is improved to a new level, and the charging efficiency is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is the overall workflow diagram of the present invention. DETAILED DESCRIPTION

[0015] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0016] See also Figure 1 As shown, this embodiment provides a control method for a charging pile device, comprising the following steps: S1. Deploy electrochemical impedance spectroscopy sensors, ultrasonic sensor arrays, and infrared thermal imaging sensor networks. Electrochemical impedance spectroscopy sensors are used to obtain data reflecting the internal electrochemical state of the battery. Ultrasonic sensor arrays are used to detect battery plate deformation, stratification, and electrolyte distribution. Infrared thermal imaging sensors are used to obtain battery surface temperature distribution. S2. Build a multimodal data fusion model based on a deep learning framework, synchronize the data collected by the electrochemical impedance spectroscopy sensor, the data collected by the ultrasonic sensor array, and the infrared thermal imaging sensor monitoring data in time and space, and fuse them at the feature level, and output a feature vector that comprehensively reflects the battery aging status; The method for spatiotemporal synchronization of multimodal data in S2 is specifically as follows: The sensor network is equipped with a clock synchronization module based on GPS timing. Each sensor has a built-in GPS receiving chip. It receives satellite signals to obtain a high-precision timestamp, which is used as the starting time benchmark for data collection to ensure that all sensors start sampling in the same absolute time coordinate system. The GPS satellite system can provide a globally unified, high-precision time standard, eliminate the clock source differences between sensors, and fundamentally ensure the consistency of data start time, avoiding synchronization errors caused by accumulated clock deviations, providing a solid time foundation for subsequent precise fusion, and greatly improving the accuracy and consistency of the system's real-time monitoring of battery status.

[0017] At a certain moment When all sensors receive a specific synchronization pulse in the GPS signal, the data acquisition circuit is triggered synchronously, and the electrochemical impedance spectroscopy data is recorded from that moment on. , Ultrasonic data And infrared thermal imaging data ,The collaborative analysis of multimodal data requires a common time reference, otherwise ,it is impossible to accurately associate the changes of different physical quantities, and provide accurate time ,correspondence for subsequent complex fusion analysis to avoid data ,mismatch.

[0018] In the post-processing stage of data collection, a time deviation correction model is established. According to the characteristics of the data transmission protocol and the network topology, the distribution law of the transmission delay from the sensor to the fusion processing center is analyzed, and the sending and receiving timestamps of the data packets are collected by embedding the time monitoring module at the key nodes of the transmission link. and , calculate the transmission delay ; This can accurately quantify the delay of data in network transmission. Different transmission paths and network loads will cause transmission delay fluctuations, which require dynamic monitoring to provide accurate delay data for subsequent precise compensation, avoid fixed compensation value errors, and ensure that the impact of transmission delay is considered during fusion processing. The least squares method is then used to fit the relationship model between transmission delay, data flow, and link load. ;in For data traffic, is the link load, To fit the coefficients, the transmission delay of each data packet is predicted and compensated. At the same time, in view of the inherent frequency drift of the sensor clock crystal, the standard clock source signal is regularly collected to calibrate the local clock deviation of the sensor. , use linear interpolation to correct the time stamp of the collected data ;in, In order to collect data timestamps, the time synchronization accuracy of multimodal data is further improved to below microseconds, the spatiotemporal consistency of data during fusion processing is ensured, the misjudgment of battery status due to time dislocation is avoided, the system's ability to capture dynamic changes in the battery is enhanced, and the reliability and timeliness of intelligent charging decisions are optimized.

[0019] The method of feature-level fusion of multimodal data in S2 is as follows: For electrochemical impedance spectroscopy data, discrete wavelet transform is used to decompose signals in different frequency bands and extract the energy characteristics of each frequency band. ; and linearly normalized to interval, highlighting the relative difference of energy at different frequencies, mining impedance data to hide aging information, and processing ultrasonic data through Hilbert-Huang transform to obtain marginal spectrum characteristics ; Standard normalization is used to highlight the changes in weak fault characteristics, and the infrared thermal imaging data is first segmented to extract the temperature mean, variance and temperature gradient characteristics of each region ; Use robust normalization to balance data distribution, enhance feature stability and noise resistance, lay a standardized data foundation for deep fusion, enable different modal features to participate in fusion at the same dimensional scale, and improve the fairness and accuracy of the fusion effect.

[0020] Construct a deep convolutional neural network fusion architecture. The input layer receives the normalized multimodal feature vector, and the convolution layer uses Convolution kernel arrangement, set 3 convolution layers to extract high-dimensional abstract features, and use the corrected linear unit as the activation function to enhance nonlinear expression. The formula is: ,in The number of convolutional layers is 1, and the pooling layer reduces the dimension and reduces the amount of calculation while retaining the key features. The attention mechanism is introduced before the fully connected layer to adaptively weight the features of each modality according to the importance of the features and enhance the contribution of key aging features. The fully connected layer integrates the features and outputs them through the Softmax function. ;in The output of the fully connected layer corresponds to The aging status score is calculated and the output is a feature vector that comprehensively reflects the probability distribution of the battery aging degree. This realizes the deep fusion and intelligent judgment of multimodal features, accurately describes the battery aging degree and state evolution trend, provides core data basis for subsequent charging strategy optimization, and improves the intelligent and refined level of battery health management of charging piles.

[0021] S3. Collect multimodal perception data of electric vehicle batteries under different charge and discharge rates, ambient temperature and humidity conditions, as well as the corresponding battery remaining life and capacity attenuation curve data. Train a battery aging prediction model based on a deep belief network. Predict the aging evolution trend of the battery during the current charging process based on real-time multimodal perception data. Based on the aging evolution trend output by the battery aging prediction model, combined with the current state of charge, temperature distribution and real-time power capacity of the charging pile, use a deep Q network algorithm to generate the optimal power adjustment strategy.

[0022] The sample preprocessing and feature optimization methods for deep belief network training in S3 are as follows: For the feature vector and collected data that comprehensively reflect the probability distribution of battery aging, the minimum-maximum normalization is used to calculate the feature vector elements. Normalize it and reduce it to , construct a matrix for all data , calculate the covariance matrix ; is the number of samples, solve for the eigenvalue With the eigenvector ,according to Arrange in descending order and select the first k principal components to construct the projection matrix , among which the cumulative contribution rate exceeds 85% Value, low-dimensional data The different dimensions of data affect the training of deep belief networks. Normalization and equalization of data, PCA dimension reduction and redundancy reduction focus on key points, improve training efficiency and stability, highlight key features of aging, avoid model overfitting and computational complexity, and create concise and efficient data input for deep belief networks.

[0023] Calculate the mutual information value between each feature and the battery remaining life and capacity decay curve data, that is: ; For feature vector elements and other data features With life data , statistical joint probability and marginal probability calculate , set the threshold ,reserve > Features are used to construct new feature sets. Since multimodal data contains redundancy, the mutual information value is the key to accurately measure features and screen for life-related correlations. Optimizing the feature set improves the learning efficiency and predictive power of the model, focuses on core aging factors to eliminate interference, and enhances the ability of deep belief networks to capture the essence of aging.

[0024] The environment modeling and strategy optimization methods of the deep Q network algorithm in S3 are as follows: Accurately determine the state space, covering the battery state of charge, temperature distribution and aging degree, and achieve accurate quantification through multi-sensor data fusion. These factors are closely intertwined and affect battery performance and charging effect, and none of them can be missing. Such a comprehensive definition builds a complete battery state perception system for the intelligent body, enabling it to have a full insight into the battery working condition.

[0025] The action space is clearly defined as the power adjustment strategy of the charging pile. According to the rated power of the battery management system and the charging protocol, the power adjustment step and the adjustable range are set. According to the real-time status of the battery, the intelligent agent selects the best action from the discrete action set for implementation. Because power directly affects the charging speed and the battery aging process, it is the core control factor. This setting provides the intelligent agent with precise and clear decision boundaries and operation options, focusing on key decision variables, flexibly and accurately outputting power, accurately balancing charging efficiency and battery life, and reducing energy loss and the risk of battery performance degradation.

[0026] Calculate the transfer probability based on massive historical charging data mining and electrochemical principle driven modeling First, the data is clustered and grouped according to battery type and environmental conditions, and the state transition pattern is mined using a deep neural network. For example, using a deep neural network, the input state ,action Prediction processed by multi-layer perceptron and activation function Probability distribution and accurate transfer probability are the core basis for intelligent agent strategy optimization, which determines the direction of decision-making, gives the intelligent agent powerful predictive ability, helps it to optimize the strategy path by anticipating the consequences of actions, and avoids the risk of adverse state transfer in advance.

[0027] The reward function is constructed by comprehensively considering multiple objectives such as charging time, battery life fluctuation, energy efficiency, etc. ,set up ;in is the rated and actual charging time, is the lifetime attenuation, is the rated life, For energy efficiency, The weights are adjusted dynamically according to the charging scenario and user needs. After repeated experiments and simulation optimization, the reward drives the intelligent agent to explore the optimization strategy, which meets the essential needs of reinforcement learning and guides the intelligent agent to independently seek the best, balance multiple objectives to achieve the optimal decision-making balance, accurately lock the best power adjustment strategy, and improve charging efficiency.

[0028] The deep Q network is subdivided into the main network and the target network. The main network parameters After random initialization, the gradient descent algorithm is used to update the optimization action evaluation value calculation capability and the target network parameters during training. Initialize to The replica is regularly copied and updated from the main network to stabilize the target Q value calculation. This division is to alleviate the problem of overestimation and fluctuation of Q value in traditional Q network training, improve learning stability, and build a robust dual-core architecture for deep Q network strategy learning. The main network explores innovation and the target network navigates stably to avoid strategy learning falling into local optimality, ensuring that strategy exploration is accurate and efficient, steadily upgraded, and accelerating convergence to the optimal power adjustment strategy.

[0029] Prioritize historical multimodal perception data according to the formula Calculate, where For sample The timing difference error, To prevent small positive numbers from being zero, ; For sample award, is the discount factor, measuring the importance of the sample, For sample The state of the next moment, For action, is the target network parameter, The current time sample status, For sample The action selected at the current moment, The target network parameters For the initialized copies, since different samples contribute greatly to strategy learning, key samples can accelerate convergence and optimization. This mechanism intelligently focuses on key samples to avoid wasting resources on low-quality sample learning, accelerates convergence to the optimal strategy, and improves the model's ability to respond quickly and make accurate decisions in complex and changeable charging scenarios, ensuring efficient and safe battery charging, and optimizing the intelligent management and control response performance of charging piles.

[0030] Sampling and replaying training according to priority, setting priority threshold , samples exceeding the threshold are included in the playback buffer area, and the probability is used during training Extract samples to update the main network parameters , using mini-batch gradient descent to update ; is the learning rate, For sample Loss function. In this process, the deep Q network is strengthened to learn, memorize and optimize strategies for key scenarios, consolidate and improve the network's decision-making accuracy and adaptability for complex and key charging scenarios, ensure the rapid and accurate generation of the optimal power strategy under extreme or complex working conditions, and ensure safe and stable operation of the battery and efficient and energy-saving charging.

[0031] The multi-model fusion prediction method for battery aging evolution trend in S3 is as follows: A long short-term memory network is constructed, whose core modules include input gate, forget gate, output gate and memory unit. The time series of multimodal perception data is input, and the output sequence features are calculated through the gating mechanism to capture dynamic changes. The deep belief network is constructed by stacking restricted Boltzmann machines. The underlying restricted Boltzmann machine receives multimodal data to learn abstract features, and optimizes weights through the contrast divergence algorithm. Since the long short-term memory network is good at processing time series data to mine dynamic features, the deep belief network can abstract feature relationships and integrate the advantages of intelligent integration to capture battery aging characteristics in all directions, enhance the model's representation of complex aging processes, and create a powerful feature extraction engine for battery aging prediction to accurately capture subtle aging trends.

[0032] The output sequence features of the long short-term memory network and the output of some hidden layers of the restricted Boltzmann machine are concatenated into a joint feature vector according to the dimension. Suppose the output dimension of the long short-term memory network is , the hidden layer dimension of the restricted Boltzmann machine , concatenating them into ;in It is represented as a restricted Boltzmann machine, and the subsequent prediction layer is trained with a joint vector. The weights are optimized according to the back-propagation algorithm. This splicing and fusion of multi-dimensional features enriches the information, synergistically improves the prediction ability, strengthens the model's understanding of aging characteristics, reduces the limitations of a single network, enables the model to learn a more comprehensive and in-depth aging pattern, and improves the prediction accuracy under complex working conditions.

[0033] Divide the data and use one-third of the data to train DBN to get the initial prediction value , calculate the residual ; For the true life value, a support vector machine regression model is constructed, and the kernel function is the radial basis function ; As parameters, the support vector machine regression model is trained according to the sequence minimum optimization algorithm to learn the residual pattern and obtain the correction function , the final predicted value ; Because DBN has errors, SVR can accurately correct the residuals to improve accuracy. The SVR is a support vector machine regression model, which can effectively correct the prediction error, enhance the overall prediction reliability, provide highly accurate aging data for intelligent control of charging piles, and improve the quality of charging services and battery health management level.

[0034] S4. Establish an electrochemical-thermal-mechanical coupling microenvironment model for each single cell in the battery pack. By arranging a network of micro temperature, pressure, and voltage sensors in the battery pack, collect the changes in the microenvironment parameters of each single cell. Use a combination of finite element analysis and data-driven methods to model and simulate the internal current distribution, temperature field distribution, and stress distribution of the single cell in real time, and adjust the charging current path and power distribution of the charging pile to each single cell.

[0035] The parameter adaptive identification method of the electrochemical-thermal-mechanical coupled microenvironment model in S4 is as follows: The extended Kalman filter (EKF) algorithm is used to build the core of the parameter adaptive identification system and construct the state equation ;in It is a vector that covers key parameters such as the concentration of electrochemical active substances, electrical conductivity, thermal conductivity, and state variables such as battery temperature, voltage, and current. are input vectors such as charging current and ambient temperature, To characterize the process noise of system random interference and model uncertainty, the battery operation is affected by multiple factors. The state equation can quantify the complex cause and effect, ensure that the model dynamically tracks the actual situation, provide a theoretical framework for system dynamic prediction, accurately capture the continuous change trajectory of the battery state, and improve the model's simulation fidelity of the battery's real-time operating conditions. It lays a solid theoretical foundation for precise charging control and helps charging piles customize charging strategies according to the "personality" of the battery.

[0036] Simultaneously construct measurement equations ; Collect value vectors for a network of micro temperature, pressure, and voltage sensors, To map state variables to measurement functions, In order to measure the noise that reflects sensor errors and environmental interference, the equation converts the complex internal state of the battery into an observable quantity. Sensor data is the key input to the model, and the measurement equation builds a bridge between the two to ensure that the data is effectively integrated into the model analysis, accurately correlating the theoretical model with the measured data, providing a direct basis for parameter identification, and enhancing the model's sensitivity to the actual battery state, thus achieving coordination between theoretical prediction and actual monitoring.

[0037] Initialize state estimate With covariance , according to the battery model, initial temperature, state of charge and empirical data or laboratory test results Initial value, It is a diagonal matrix, and the diagonal element values ​​are estimated according to the parameter measurement accuracy and uncertainty, giving the model initial state cognition and uncertainty quantitative description. This reasonable initial value and covariance initialization help the EKF algorithm to quickly converge to the true parameter value, avoid initial iteration oscillation or divergence of the algorithm, accelerate the parameter identification process, ensure that the model is accurate to the actual state of the battery from the beginning, and improve the overall calculation efficiency and parameter identification timeliness.

[0038] For each sampling time k, the prediction step calculates the prior estimate based on the state equation ; This formula represents the current moment The state of the previous moment is predicted. State estimation and control input , through the state transfer function Calculate the prior state estimate ; Predict the next state according to the system dynamic model, provide a reference benchmark for the update, and calculate the covariance , which calculates the prior state covariance matrix , used to quantify the uncertainty of the predicted state, where is the state transition matrix, defined as , that is, the state transfer function The partial derivatives of the state variables describe the effect of state changes on error propagation. It was the last moment The state estimation covariance matrix describes the uncertainty of the state estimation, It is the process noise covariance matrix, which describes the random error of the system during the state transfer process. This prediction is based on the current state and the model to predict the future, prepare for the subsequent measurement data fusion, improve the forward-looking nature of parameter estimation, reduce the impact of measurement lag, gain early insight into the trend of battery state changes, enhance the model's ability to respond to sudden changes in operating conditions, reserve time margin for pre-adjustment of charging strategies, and improve the control stability of charging piles in dynamic environments.

[0039] Update step based on measured value Calculate Kalman gain ,in, is the observation matrix, describing the measured values With status The relationship between is the measurement noise covariance, describing the uncertainty of the measurement equipment, Represents the total uncertainty of the measurement, including the uncertainty of the prediction and measurement error , the function of this formula is to calculate a weight factor , which is used to balance the prediction and measurement information to make the final estimate as accurate as possible, and then update the state estimate , which uses the measured value To correct the prior state estimate , and obtain the updated state estimate ,in, represents the error between the measured value and the predicted value (i.e., the observation residual), By adjusting the weights of the observation residuals, the state estimate is made as close to the true value as possible.

[0040] Finally, the fused measurements correct the prediction bias and update the covariance ,in, Represents the factor to correct the a priori uncertainty, multiplied by After that, the prediction error is reduced and the estimation uncertainty is adjusted. This update integrates the measurement data to correct the prediction error, optimize the parameter estimation, dynamically calibrate the parameters in real time, accurately track the changes in battery parameters, improve the model's adaptability to battery aging and operating condition fluctuations, and ensure that the charging strategy of the charging pile is optimized in real time with the battery status.

[0041] The modeling and simulation control methods of finite element analysis and data-driven integration in S4 are as follows: A three-dimensional solid model is constructed based on the precise geometric structure and material properties of the battery. When dividing the grid, fine tetrahedral grids are used for key parts of the battery, such as the electrode-electrolyte interface and areas with large temperature gradients, and other parts are moderately coarsened, taking into account both calculation accuracy and resource consumption, setting control equations, heat conduction equations ; describes the heat transfer, where The heat capacity is determined by the density of the material. Specific heat capacity affects the rate of temperature change. Thermal conductivity determines how fast heat diffuses. The heat source term includes electrochemical reaction heat, electrical conduction equation ; describes the current conduction path, is the conductivity, controlling the current, Expressed as potential, describing the battery charge and discharge current distribution, mechanical equilibrium equation ; Used to control the internal stress balance of the battery, Including the battery's own weight and internal pressure, defining boundary conditions, such as the precise current density on the electrode surface is set according to the charging strategy, and the thermal convection heat dissipation conditions of the battery shell are determined according to the ambient temperature and heat dissipation coefficient. The accurate physical model needs to comprehensively consider the structure, characteristics, physical processes and boundary interactions to provide a basis for subsequent precise simulation and regulation.

[0042] Comprehensively collect battery microenvironment parameters and charge and discharge performance data under different charge and discharge rates, ambient temperature and humidity, and battery aging degree, and build an artificial neural network. The input layer receives sensor data and charge and discharge rate information, and after hidden layer neuron processing, the output layer outputs the temperature prediction value of the temperature field node. The network weights are optimized by back propagation algorithm training based on large historical experience data to minimize the prediction error. ; is the true temperature value, The network predicted values ​​are used to calibrate the finite element model parameters, such as adjusting the thermal conductivity of the material according to the predicted temperature deviation. Due to the actual battery manufacturing tolerance, material aging and operating condition complexity, the finite element model has deviations. Data-driven calibration can effectively correct them to ensure that the simulation is closely aligned with the actual operating conditions.

[0043] During the charging process, the model input is updated at fixed time intervals based on real-time sensor data collection, finite element equations and data-driven model predictions are solved in real time, and dynamic current, temperature, and stress distributions are quickly obtained with the help of a high-performance computing platform to construct a multi-objective optimization function, such as ;in, is the maximum temperature change, is the maximum current change, is the maximum stress change, is the rated value, As weight, the genetic algorithm is used to initialize the population to randomly generate chromosome encoding of charging current path and power allocation strategy, evaluate the individual merits according to the fitness function, and iterate through selection, crossover and mutation operations to calculate the optimal charging current path and power allocation strategy. Minimize and meet battery safety constraints at the same time. Such real-time regulation optimizes charging according to the battery, realizes precise and intelligent regulation of the entire charging process, improves battery performance consistency, and enhances the charging pile's ability to adapt to complex working conditions.

[0044] The control method of the above-mentioned charging pile equipment, by constructing a multimodal perception data fusion model, integrating electrochemical impedance spectroscopy, ultrasonic and infrared thermal imaging sensor data, deeply mining the battery aging status, and predicting the aging trend with the help of a deep belief network, generates a dynamic optimal power strategy based on a deep Q network algorithm, improves charging efficiency, and increases battery life. For single-cell battery management, relying on an electrochemical-thermal-mechanical coupling microenvironment model and a finite element and data-driven fusion strategy, the single-cell battery microenvironment is closely monitored and finely balanced and regulated, and the temperature and voltage differences between cells are strictly controlled within an extremely small range, thereby greatly improving the performance consistency and stability of the battery pack and improving the charging efficiency of single-cell batteries of electric vehicles.

[0045] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A control method for a charging pile device, characterized in that: The following steps are involved: S1. Deploy electrochemical impedance spectroscopy sensors, ultrasonic sensor arrays, and infrared thermal imaging sensor networks. Electrochemical impedance spectroscopy sensors are used to obtain data reflecting the internal electrochemical state of the battery. Ultrasonic sensor arrays are used to detect battery plate deformation, stratification, and electrolyte distribution. Infrared thermal imaging sensors are used to obtain battery surface temperature distribution. S2. Build a multimodal data fusion model based on a deep learning framework, synchronize the data collected by the electrochemical impedance spectroscopy sensor, the data collected by the ultrasonic sensor array, and the infrared thermal imaging sensor monitoring data in time and space, and fuse them at the feature level, and output a feature vector that comprehensively reflects the battery aging status; S3. Collect multimodal sensing data of electric vehicle batteries under different charge and discharge rates, ambient temperature and humidity conditions and the corresponding battery remaining life and capacity decay curve data, train a battery aging prediction model based on a deep belief network, predict the aging evolution trend of the battery during the current charging process based on real-time multimodal sensing data, and generate the optimal power adjustment strategy using a deep Q network algorithm based on the aging evolution trend output by the battery aging prediction model, combined with the current state of charge, temperature distribution and real-time power capacity of the charging pile; S4. Establish an electrochemical-thermal-mechanical coupling microenvironment model for each single cell in the battery pack. By arranging a network of micro temperature, pressure, and voltage sensors in the battery pack, collect the changes in the microenvironment parameters of each single cell. Use a combination of finite element analysis and data-driven methods to model and simulate the internal current distribution, temperature field distribution, and stress distribution of the single cell in real time, and adjust the charging current path and power distribution of the charging pile to each single cell.

2. A control method for a charging pile device according to claim 1, characterized in that: The method for spatiotemporal synchronization of multimodal data in S2 is specifically as follows: The sensor network is equipped with a clock synchronization module based on GPS timing. Each sensor has a built-in GPS receiving chip to receive satellite signals to obtain a timestamp, which is used as the starting time reference for data collection. When all sensors receive a specific synchronization pulse in the GPS signal, the data acquisition circuit is triggered synchronously to record electrochemical impedance spectroscopy data, ultrasonic data, and infrared thermal imaging data from that moment on. After completing data collection, a time deviation correction model is established to analyze the transmission delay of data from the sensor to the fusion processing center. By embedding a time monitoring module in the key nodes of the transmission link, the sending and receiving timestamps of the data packets are collected, the transmission delay is calculated, and the least squares method is used to fit the relationship model between the transmission delay and the data flow and link load, and the transmission delay of each data packet is predicted and compensated. At the same time, the standard clock source signal is regularly collected to calibrate the local clock deviation of the sensor, and the timestamp of the collected data is corrected by linear interpolation.

3. A control method for a charging pile device according to claim 2, characterized in that: The method for feature-level fusion of multimodal data in S2 is as follows: The electrochemical impedance spectroscopy data, ultrasonic data and infrared thermal imaging data were normalized to obtain the energy characteristics of each frequency band, marginal spectrum characteristics, and the temperature mean, variance and temperature gradient characteristics of each region; A deep convolutional neural network fusion architecture is constructed, which includes an input layer, a convolutional layer, a pooling layer and a fully connected layer. The input layer receives a normalized multimodal feature vector, the convolutional layer is used to extract features, the pooling layer is used to reduce the dimension of the extracted features, and the fully connected layer is used to output a feature vector that comprehensively reflects the probability distribution of battery aging.

4. A control method for a charging pile device according to claim 3, characterized in that: The sample preprocessing and feature optimization method for deep belief network training in S3 is as follows: Normalize the eigenvector that comprehensively reflects the probability distribution of battery aging degree with the collected multimodal sensing data and battery life-related data, calculate the data covariance matrix and solve the eigenvalues ​​and eigenvectors, select the main components according to the eigenvalues, and project the high-dimensional data into a low-dimensional space; The mutual information value between each feature and the remaining battery life and capacity decay curve data is calculated, and the key features are selected according to the mutual information value. The mutual information value is used to measure the correlation between the feature and the target variable.

5. A control method for a charging pile device according to claim 4, characterized in that: The environment modeling and strategy optimization method of the deep Q network algorithm in S3 is as follows: The battery charging process is abstracted into a Markov decision process environment model, and the state space, action space and transition probability are defined. The state space includes the battery state of charge, temperature distribution and aging degree. The action space is the charging pile power adjustment strategy. The transition probability is used to describe the probability of transferring to another state after performing an action in a certain state. Then, the reward function is obtained based on historical data and model calculation. The deep Q network is divided into a main network and a target network. The main network is used to select action evaluation values, and the target network is used to calculate target Q values. Priorities are then set for historical multimodal perception data. The priorities are calculated by a preset fixed formula. Sample data whose priorities exceed the preset priority threshold are replayed for training to complete the optimization.

6. A control method for a charging pile device according to claim 5, characterized in that: The multi-model fusion prediction method for battery aging evolution trend in S3 is as follows: Fusion of long short-term memory network and deep belief network, wherein the long short-term memory network is used to process multimodal perception data time series to capture dynamic features, and the deep belief network is used to learn data abstract features and aging relationships; During fusion, the output sequence features of the long short-term memory network and some hidden layer outputs of the deep belief network are spliced ​​into a joint feature vector and input into the subsequent layer. Then, based on the deep belief network's prediction of the aging trend, the support vector machine regression model is introduced to learn the prediction residual correction. The deep belief network is trained with one-third of the data to obtain the preliminary prediction value, and the residual is calculated. Finally, the support vector machine regression model is trained with the residual to obtain the correction function, and the final prediction value is obtained based on this.

7. The control method of a charging pile device according to claim 1, characterized in that: The parameter adaptive identification method of the electrochemical-thermal-mechanical coupled microenvironment model in S4 is as follows: The extended Kalman filter algorithm is used to identify the key parameters of the model in real time, and the state equation and measurement equation are constructed, and the state estimation and covariance are initialized. The state equation is used to predict the state of the system at the next moment, and the measurement equation is used to describe the relationship between the sensor and the system state. At each sampling moment, the prediction step calculates the prior estimate and covariance, and the update step calculates the Kalman gain based on the measurement value and updates the state estimate and covariance.

8. The control method of a charging pile device according to claim 1, characterized in that: The modeling and simulation control method of the finite element analysis and data-driven integration in S4 is as follows: Constructing a three-dimensional solid model according to the battery geometry and material properties, dividing the grid, and setting control equations, wherein the control equations include heat conduction equations, electrical conduction equations, and mechanical equilibrium equations; Redefine the boundary conditions, collect the battery microenvironment parameters and charge and discharge performance data under multiple working conditions, the microenvironment parameters include temperature, pressure and voltage, build an artificial neural network, the input parameters are sensor data, the output is the temperature prediction value of the temperature field node, the prediction value is used to calibrate the finite element model parameters; During the charging process, the model input is updated based on the real-time collected sensor data, the finite element equations and data-driven model predictions are solved in real time, the dynamic current, temperature, and stress distributions are obtained, a multi-objective optimization function is constructed, and a genetic algorithm is used to calculate the optimal charging current path and power allocation strategy, so that the multi-objective optimization function is minimized while satisfying the battery safety constraints.

Images