A charging control method for electric motorcycle

Abstract

The present application provides an electric motorcycle charging control method based on battery power, comprising: when the electric motorcycle is charging, obtaining the tilt angle data of the electric motorcycle side bracket, if the tilt angle is within a preset range, judging that the vehicle body is in a tilt state when charging, and triggering a monitoring program for the flow state of the battery pack coolant; real-time acquisition of the temperature data of the battery cells in the identified coolant uneven distribution area through a temperature sensor array, performing gradient analysis on the temperature data, determining the location and severity of the thermal imbalance fault, and providing data support for subsequent charging strategy adjustment and cooling system optimization; simulating the temperature distribution and flow trajectory of the coolant at different tilt angles, and determining the heat dissipation risk area at the current tilt angle by comparing and analyzing the uniformity of the temperature distribution and the smoothness of the flow.

Description

Technical Field

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

[0002] The compact structural design of electric motorcycles results in a small space between the battery, controller and motor, making heat dissipation difficult. When the vehicle body is tilted for charging, the coolant inside the battery pack is unevenly distributed, and the temperature of the battery cells on the side close to the ground rises much faster than the top, increasing the risk of hot spots. In a high temperature environment, the battery temperature rises rapidly, affecting charging efficiency and safety. The unique side stand charging design of electric motorcycles allows the vehicle body to tilt at an angle of 15-20 degrees, further exacerbating the temperature gradient inside the battery pack, with the temperature difference between the top and bottom reaching 20-30%. The combination of the compact structural layout and the tilted charging design of the vehicle body poses severe challenges to the heat dissipation of electric motorcycles, and a targeted thermal management optimization solution is urgently needed. Summary of the invention

[0003] The present invention provides an electric motorcycle charging control method, which mainly includes:

[0004] When the electric motorcycle is charging, the tilt angle data of the side stand of the electric motorcycle is obtained. If the tilt angle is within a preset range, it is determined that the vehicle body is in a tilted state when charging, and a monitoring program for the flow state of the battery pack coolant is triggered;

[0005] The flow sensor installed inside the battery pack collects the flow rate data of the coolant in different areas in real time. Based on the preset flow rate threshold, it determines whether there is an area with uneven coolant flow and identifies high-risk areas with uneven coolant distribution. At the same time, the battery management system obtains the power data of each battery cell in the battery pack.

[0006] The temperature sensor array collects the temperature data of the battery cells in the area with uneven coolant distribution in real time, performs gradient analysis on the temperature data, determines the location and severity of the thermal imbalance fault, and provides data support for subsequent charging strategy adjustment and cooling system optimization.

[0007] According to the location and severity of the thermal imbalance fault, the charging strategy and cooling system operating parameters are dynamically adjusted. For battery cells with severe thermal imbalance faults and low power, the strategy of reducing charging current and enhancing local cooling is adopted. For battery cells with severe thermal imbalance faults and high power, the strategy of extending charging time and enhancing local cooling is adopted.

[0008] Simulate the temperature distribution and flow trajectory of the coolant at different tilt angles, and determine the heat dissipation risk area at the current tilt angle by comparing and analyzing the uniformity of temperature distribution and smoothness of flow;

[0009] For heat dissipation risk areas, control the speed and flow of the coolant pump to enhance local heat dissipation, and / or control the speed of the electronic fan on the battery pack housing to enhance air convection, and / or control the active thermal management system to balance the temperature distribution inside the battery pack, and temporarily reduce the charge and discharge power of the battery cells in the heat dissipation risk area when the heat dissipation risk is greater than the preset safety threshold;

[0010] A battery life prediction model is established based on data from battery charge, temperature, charging process control, and cooling system optimization. The battery's health status and remaining life are evaluated through a machine learning algorithm. The location and severity of thermal imbalance failures are included in the evaluation indicators, and a battery pack maintenance and replacement strategy is formulated to ensure the safe and reliable operation of electric motorcycles.

[0011] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0012] The present invention discloses a charging control method for an electric motorcycle. By detecting the tilt angle of the vehicle body, coolant flow monitoring is triggered within the range of 15-20 degrees. A flow sensor is used to collect coolant flow rate data in real time to identify unevenly distributed areas. Combined with the power data obtained by the battery management system and the temperature data collected by the temperature sensor array, a gradient analysis is performed to determine the location and severity of the thermal imbalance fault. According to the analysis results, the charging strategy and cooling system parameters are dynamically adjusted to reduce the charging current and strengthen local cooling for monomers with severe thermal imbalance and low power, and to extend the charging time and strengthen cooling for monomers with high power. At the same time, the temperature distribution and flow trajectory at different tilt angles are simulated to determine the heat dissipation risk area. For the risk area, the heat dissipation is enhanced by controlling the coolant pump, the electronic fan and the active thermal management system, and the charging and discharging power is reduced when necessary. Finally, a battery life prediction model is established based on the collected data, the health status is evaluated and a maintenance strategy is formulated to ensure the safe and reliable operation of the electric motorcycle. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 The present invention is a flow chart of an electric motorcycle charging control method.

[0014] Figure 2 It is a schematic diagram of a charging control method for an electric motorcycle of the present invention.

[0015] Figure 3 It is another schematic diagram of an electric motorcycle charging control method of the present invention. DETAILED DESCRIPTION

[0016] The following will describe the technical solutions in the embodiments of the present invention in detail in conjunction with the accompanying drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention.

[0017] like Figure 1-3 In this embodiment, a method for controlling charging of an electric motorcycle may specifically include:

[0018] Step S101, when the electric motorcycle is charging, the tilt angle data of the side bracket of the electric motorcycle is obtained. If the tilt angle is within a preset range, it is determined that the vehicle body is in a tilted state when charging, and the monitoring program of the flow state of the battery pack coolant is triggered.

[0019] The original data collected by the three-axis acceleration sensor installed on the main control panel of the vehicle body is obtained, and the inclination angle value of the side bracket of the electric motorcycle is calculated by the cosine theorem, and the inclination angle value is transmitted to the vehicle control unit. According to the inclination angle value received by the vehicle control unit, it is compared with the preset angle range. If the inclination angle value falls within the preset angle range, it is determined that the vehicle body is in an inclined state under the charging state. After determining that the vehicle body is in an inclined state under the charging state, the battery pack coolant flow state monitoring program is started, and the real-time flow data is collected by the flow sensor installed on the coolant pipeline, combined with the battery temperature information obtained by the temperature sensor installed on the battery pack shell. The coolant flow data and battery temperature data are obtained, and the support vector regression algorithm is used to establish a cooling effect evaluation model, and the coolant flow and battery temperature are used as input features to obtain the cooling effect score. According to the cooling effect score output by the cooling effect evaluation model, the speed of the coolant pump is controlled by a pulse width modulation signal to achieve precise adjustment of the coolant flow rate, so as to optimize the battery temperature control during the charging process.

[0020] Specifically, the tilt angle sensor data of the side bracket of the electric motorcycle is obtained, the original data is collected by the three-axis acceleration sensor installed on the main control panel of the vehicle body, the accurate tilt angle value is calculated by the cosine theorem, and the angle value is transmitted to the vehicle control unit for subsequent judgment. After receiving the tilt angle value, the vehicle control unit compares it with the preset angle range of 15-20 degrees. If the tilt angle value falls within this range, it is determined that the vehicle body is in a tilted state under charging, and the subsequent monitoring program execution instruction is triggered. According to the trigger instruction, the battery pack coolant flow state monitoring program is started, and the real-time flow data is collected by the flow sensor installed on the coolant pipeline. Combined with the battery temperature information obtained by the temperature sensor installed on the battery pack shell, the collected data is stored in the vehicle data storage unit. The coolant flow data and battery temperature data are read from the vehicle data storage unit, and the support vector regression algorithm is used to establish a cooling effect evaluation model, with the coolant flow and battery temperature as input features, and the cooling effect score is output. According to the cooling effect score output by the model, the speed of the coolant pump is controlled by the pulse width modulation signal to achieve accurate adjustment of the coolant flow, thereby optimizing the battery temperature control during the charging process.

[0021] The tilt angle of the electric motorcycle side bracket is collected by the MPU-6050 three-axis acceleration sensor installed on the main control board of the vehicle body. The sensor has a measurement range of ±2g and a resolution of 16 bits. The cosine theorem is used to calculate the tilt angle. Let x, y, and z be the three-axis acceleration values, and the tilt angle θ=arccos(z / √(x^2+y^2+z^2)). The calculated angle value is de-noised by the Kalman filter algorithm, and the filtered angle value is transmitted to the vehicle control unit STM32F407 chip through the CAN bus for judgment. After receiving the tilt angle value, the vehicle control unit compares it with the preset 15-20 degree angle range, and uses the interval judgment algorithm. If 15°≤θ≤20°, it is determined that the vehicle body is in a tilted state under charging, and the monitoring program is triggered to execute instructions. After the monitoring program is started, the real-time flow data is collected by the turbine flow sensor installed on the coolant pipeline. The sensor range is 0.5-10L / min and the accuracy is ±1%. At the same time, the PT100 platinum resistance temperature sensor installed on the battery pack shell is used to obtain battery temperature information, with a temperature measurement range of -50℃ to 150℃ and an accuracy of ±0.1℃. The collected data is stored in the on-board 64GB solid-state hard drive at a frequency of 10Hz to form a time series database. The on-board computing unit reads the coolant flow and battery temperature data of the last 30 minutes from the database, and uses the SVR (Support Vector Regression) algorithm in the scikit-learn library to establish a cooling effect evaluation model. The model uses coolant flow and battery temperature as input features and outputs a cooling effect score of 0-100. According to the score output by the model, a 20kHz PWM signal is generated by the STM32F407 chip to control the speed of the coolant pump. The PWM duty cycle is linearly related to the score. For every 10 points increase in the score, the duty cycle increases by 5%, achieving precise adjustment of the coolant flow, thereby optimizing battery temperature control while ensuring charging efficiency.

[0022] Step S102, through the flow sensor installed inside the battery pack, the flow rate data of the coolant in different areas is collected in real time, and according to the preset flow rate threshold, it is determined whether there is an area with uneven coolant flow, and the high-risk area of ​​uneven coolant distribution is identified. At the same time, the power data of each battery cell in the battery pack is obtained through the battery management system.

[0023] The coolant flow rate data collected by multiple thermal flow sensors inside the battery pack are obtained; the flow rate data is compared with a preset flow rate threshold; if the flow rate in a certain area is lower than the threshold, it is determined that there is uneven coolant flow in the area; a clustering algorithm is used to group the areas with uneven flow to obtain high-risk areas with uneven coolant distribution; the power data of each battery cell in the battery pack is obtained from the battery management system through the communication interface, the power data is correlated with the coolant flow rate data, and a decision tree algorithm is used to establish a relationship model between the battery cell power and the cooling demand; according to the cooling demand prediction value output by the relationship model, the increase or decrease ratio of the coolant flow required for each area is calculated, the ball valve in the battery pack cooling system is adjusted, the coolant flow is increased for high-risk areas, and the flow in other areas is reduced at the same time, so as to achieve dynamic balanced distribution of coolant inside the battery pack.

[0024] Specifically, multiple FT-210 thermal flow sensors with a measurement range of 0.1-2m / s are installed inside the battery pack. The sensors are arranged in different cooling circuits according to the structural characteristics of the battery pack. The STM32F407 microcontroller is used as a data acquisition module to obtain the coolant flow rate data of each area in real time, and the collected data is transmitted to the RaspberryPi4 as the central processing unit for subsequent analysis. After receiving the flow rate data, the central processing unit compares it with the flow rate threshold of 0.5m / s pre-set according to the battery pack design specifications and thermal simulation results. If the flow rate in a certain area is lower than the threshold, it is determined that there is uneven coolant flow in the area. The K-means clustering algorithm is used to group the areas with abnormal flow rates and identify high-risk areas with uneven coolant distribution. The power data of each battery cell in the battery pack is obtained from the battery management system through the communication interface, and the power data is correlated with the coolant flow rate data. The C4.5 decision tree algorithm is used to establish a relationship model between the battery cell power and cooling requirements. According to the cooling demand forecast value output by the relational model, the increase and decrease ratio of the coolant flow required for each area is calculated, and the ball valve driven by a stepper motor in the battery pack cooling system is adjusted with a control accuracy of 0.1%. The coolant flow is increased for high-risk areas, while the flow in other areas is reduced to achieve dynamic balance distribution of coolant inside the battery pack and improve the overall cooling efficiency. The FT-210 thermal flow sensor is installed in the cooling circuit inside the battery pack. Three sensors are installed in each cooling circuit, and a total of 18 sensors cover the entire battery pack. The STM32F407 microcontroller collects flow rate data at a sampling frequency of 100Hz and transmits the data to the RaspberryPi4 central processing unit through the CAN bus. After receiving the data, the central processing unit first performs data preprocessing, including removing outliers and smoothing. The preprocessed data is compared with the flow rate threshold of 0.5m / s, and the area below the threshold is marked as abnormal. The K-means clustering algorithm uses Euclidean distance as the distance metric, sets the number of clusters K=3, and groups the abnormal areas. The clustering results are used to identify high-risk areas, and the risk level is sorted according to the difference between the cluster center and the threshold. At the same time, the power data of each battery cell is obtained from the battery management system through the CAN bus, with a sampling frequency of 1Hz. The C4.5 decision tree algorithm uses the information gain ratio as the feature selection criterion, takes the battery cell power, ambient temperature and historical cooling demand as input features, and cooling demand as the target variable to build a decision tree model with a depth of 5. The model training uses the historical data of the last 7 days and is updated every 24 hours. According to the output of the decision tree model, the cooling demand forecast value of each area is calculated and compared with the current flow rate to obtain the flow adjustment ratio. For example, if the cooling demand in a certain area is predicted to increase by 20% and the current flow rate is 0.4m / s, the target flow rate is 0.48m / s.The ball valve driven by the stepper motor is adjusted accordingly according to the flow adjustment ratio, with each adjustment step being 0.1%. Precise control is achieved through the PID control algorithm, and the adjustment cycle is 10 seconds.

[0025] Step S103, collecting temperature data of battery cells in the identified coolant uneven distribution area in real time through the temperature sensor array, performing gradient analysis on the temperature data, determining the location and severity of the thermal imbalance fault, and providing data support for subsequent charging strategy adjustment and cooling system optimization.

[0026] Acquire temperature data of the area where the coolant is unevenly distributed; pre-process the temperature data, including smoothing the temperature data using a sliding window method; calculate the mean and standard deviation of the temperature data according to the window size and step size; remove abnormal data points to obtain a smoothed temperature data sequence. Determine the thermal imbalance fault area, including: calculate the temperature gradient using a finite difference method; if the temperature gradient exceeds a preset threshold, it is determined to be a thermal imbalance fault area; determine the severity of the fault according to the size of the temperature gradient. Identify thermal imbalance fault modes, including: use convolutional neural networks to perform deep learning on temperature data; perform supervised learning through historical fault data; combine the output of temperature gradient analysis to identify hot spots, hot zones, and local overheating thermal imbalance fault modes. Generate a thermal imbalance fault report, including: determine the location and severity of the thermal imbalance fault according to the thermal imbalance fault mode and fault severity; adjust the charging current and coolant flow distribution according to the thermal imbalance fault report.

[0027] Specifically, a K-type thermocouple temperature sensor array is installed in the identified area with uneven coolant distribution, with a measurement range of -200℃ to 1350℃ and an accuracy of ±0.5℃. A matrix layout is adopted, and 4 sensors are arranged on the surface of each battery cell. The data acquisition module using the STM32F103 microcontroller as the core processor acquires temperature data in real time at a sampling frequency of 10Hz, and transmits the collected temperature data to the central processing unit for subsequent processing. After the central processing unit receives the temperature data, it uses the sliding window method to pre-process the data. The window size is set to 10 seconds and the step size is 1 second. The weighted average method is used to calculate the temperature value in the window. By calculating the average value and standard deviation in the window, the abnormal data points are eliminated to obtain a smoothed temperature data sequence, which provides a reliable data basis for subsequent gradient analysis. The preprocessed temperature data is subjected to gradient analysis, and the temperature gradient is calculated using the finite difference method in the central difference format. The area with the fastest temperature change is determined by comparing the temperature difference between adjacent sensors. If the temperature gradient exceeds the preset threshold of 5°C / cm set according to the battery pack design specifications and safety requirements, it is determined as a thermal imbalance fault area, and the severity of the fault is determined according to the magnitude of the temperature gradient. A convolutional neural network with 5 convolutional layers and 3 fully connected layers is used to perform deep learning on the temperature data. Supervised learning is performed through historical fault data to identify different types of thermal imbalance fault modes, such as hot spots, hot spots, and local overheating. Combined with the output of the temperature gradient analysis, a report on the location and severity of the thermal imbalance fault is generated. The charging current and coolant flow distribution are dynamically adjusted according to the report to achieve charging strategy adjustment and cooling system optimization. The K-type thermocouple temperature sensor array is arranged in a 5cm×5cm grid in the battery pack, and the four sensors on the surface of each battery cell are located in the four corners. The STM32F103 microcontroller collects temperature data at a frequency of 10Hz through a 16-bit ADC with a resolution of 0.1°C. The collected raw data is processed by a sliding window with a size of 10 seconds and a step length of 1 second. The temperature value in the window is calculated by the Gaussian weighted average method, and the weight coefficient is [0.1, 0.2, 0.4, 0.2, 0.1]. The criterion for judging abnormal data points is more than 3 times the standard deviation. The temperature gradient calculation formula in the central difference format is (T(i+1)-T(i-1)) / (2*dx), where dx is the sensor spacing, T(i+1) represents the temperature value at a spacing dx to the right of a point i in space, and T(i-1) represents the temperature value at a spacing dx to the left of a point i in space. If the calculated temperature gradient exceeds 5℃ / cm, it is marked as a thermal imbalance area, and the severity is divided into 3 levels: 5-7℃ / cm is mild, 7-10℃ / cm is moderate, and >10℃ / cm is severe.The convolutional neural network structure contains 5 convolutional layers, with 32, 64, 128, 256, and 512 3×3 convolution kernels and 3 fully connected layers, with 1024, 512, and 256 neurons, respectively, using ReLU activation function and batch normalization. The network training uses the Adam optimizer, with a learning rate of 0.001, a batch size of 64, and 100 training rounds. According to the generated thermal imbalance fault report, the charging current adjustment range is 0.1C-1C, with a step size of 0.1C; the coolant flow adjustment range is 0.5-2L / min, with a step size of 0.1L / min. The charging strategy and cooling system parameters are updated every 30 seconds to continuously optimize the temperature balance of the battery pack.

[0028] Step S104, dynamically adjust the charging strategy and cooling system operating parameters according to the location and severity of the thermal imbalance fault. For battery cells with severe thermal imbalance faults and low power, adopt a strategy of reducing charging current and enhancing local cooling. For battery cells with severe thermal imbalance faults and high power, adopt a strategy of extending charging time and enhancing local cooling.

[0029] The thermal imbalance fault location and severity information of the battery cell are obtained, and the real-time power data of the battery cell is read from the battery management system. According to the thermal imbalance fault information and power data, a filtering algorithm is used for data fusion to obtain a comprehensive data report containing the battery ID, fault location coordinates, severity level and current power percentage. According to the comprehensive data report, the charging strategy is dynamically adjusted by a fuzzy logic controller, the controller input is the fault severity and power level, and the output is the charging current adjustment value and the charging time adjustment coefficient. If the thermal imbalance fault of the battery cell is serious and the power is low, the charging current is reduced, and the charging current adjustment range is from the preset minimum value to the maximum value. For battery cells with severe thermal imbalance faults and high power, the optimal charging time is calculated, and the range of extending the charging time is from the preset minimum multiple to the maximum multiple of the original planned charging time. After receiving the local cooling enhancement instruction, the fuzzy adaptive PID controller is called to dynamically adjust the coolant flow and temperature according to the location and severity of the thermal imbalance fault, and the PID parameters are adjusted in real time according to the temperature deviation and change rate.

[0030] Specifically, the location and severity information of the thermal imbalance fault is obtained, and the real-time power data of each battery cell is read from the battery management system. The data is fused through the Kalman filter algorithm to generate a comprehensive data report containing the battery ID, fault location coordinates, severity level (level 1-5) and current power percentage. According to the comprehensive data report, a fuzzy logic controller is used to dynamically adjust the charging strategy. The controller input is the fault severity and power level, and the output is the charging current adjustment value and the charging time adjustment coefficient. The power level is divided into three levels: low 0-30%, medium 30-70%, and high 70-100%. If the thermal imbalance fault of the battery cell is severe 4-5 and the power is low 0-30%, the charging current is reduced. The charging current adjustment range is 0.1C to 0.5C, and the adjustment step is 0.1C. For battery cells with severe thermal imbalance faults of level 4-5 and high power of 70-100%, the optimal charging time is calculated through an algorithm based on model predictive control. The charging time is extended from 1.1 to 1.5 times the original planned charging time, with a step length of 0.1 times. After receiving the local cooling enhancement instruction, the cooling system control unit calls the fuzzy adaptive PID controller to dynamically adjust the coolant flow and temperature according to the location and severity of the thermal imbalance fault. The flow adjustment range is 0.5L / min to 2L / min, and the temperature adjustment range is 15℃ to 25℃. The PID parameters are adjusted in real time according to the temperature deviation and change rate to achieve accurate local cooling control. The Kalman filtering algorithm uses a state space model. The state vector contains the fault location coordinates (x, y), severity S and power percentage E, and the observation vector contains sensor measurements and battery management system readings. The state transfer matrix is ​​the identity matrix, the process noise covariance Q = diag(0.1, 0.1, 0.05, 0.01), and the observation noise covariance R = diag(0.2, 0.2, 0.1, 0.05). The state estimate is updated every 100ms to generate a comprehensive data report. The fuzzy logic controller uses 5 input membership functions (very low, low, medium, high, very high) and 5 output membership functions (significantly reduced, slightly reduced, maintained, slightly increased, and significantly increased), and adopts the Mamdani inference rule. For example, if the fault severity is level 4 and the power is 20%, the output is "significantly reduced" charging current, and the corresponding adjustment value is -0.3C. The optimal charging time calculation based on model predictive control uses a battery equivalent circuit model, with a prediction time domain of 30 minutes and a control time domain of 5 minutes. The objective function comprehensively considers the charging speed and temperature rise. If the predicted temperature rise exceeds 5℃ / hour, the charging time is extended by 1.2 times. The fuzzy rule base of the fuzzy adaptive PID controller contains 49 rules, with the input being the temperature deviation e and the deviation change rate ec, and the output being the PID parameter increments ΔKp, ΔKi, and ΔKd. For example, when e is "positive and large" and ec is "negative and small", ΔKp is "positive and medium", ΔKi is "zero", and ΔKd is "negative and small".The frequency of real-time PID parameter adjustment is 10Hz, ensuring that the cooling system responds quickly to thermal imbalance. The charging strategy and cooling parameters are updated every 5 seconds to continuously optimize the battery pack temperature balance and charging efficiency.

[0031] Step S105, simulating the temperature distribution and flow trajectory of the coolant at different tilt angles, and determining the heat dissipation risk area at the current tilt angle by comparing and analyzing the uniformity of the temperature distribution and the smoothness of the flow.

[0032] The actual size parameters of the battery pack, flow channel and radiator are obtained, and a three-dimensional geometric model is constructed according to the parameters. The coolant flow channel, battery cell position and radiator position are set for the three-dimensional geometric model to obtain a complete battery cooling system model. The battery cooling system model is numerically simulated at different inclination angles using ANSYS Fluent software, and the Navier-Stokes equations and energy equation partial differential equations containing velocity, pressure and temperature are solved to obtain the velocity field and temperature field distribution of the coolant. According to the temperature field distribution, the Shannon entropy is used to calculate the uniformity index of the temperature distribution. The uniformity index is obtained by discretizing the temperature field into n intervals, calculating the probability pi of each interval, and obtaining the entropy value. Calculate the resistance coefficient of the coolant flow, judge the patency of the coolant flow, and perform weighted summation of the uniformity index and the patency index to obtain a comprehensive evaluation index. Based on the support vector machine algorithm, the tilt angle and the comprehensive evaluation index are used as input features to train a heat dissipation risk prediction model. The model optimizes the penalty parameter C and the kernel parameter γ through the radial basis function kernel and the grid search method to predict and identify the heat dissipation risk area under the current tilt angle.

[0033] Specifically, SolidWorks software was used to construct accurate three-dimensional geometric models of the battery pack, flow channel and radiator according to the actual size, and the coolant flow channel, battery cell position and radiator position were set. ANSYS Fluent software was used to perform numerical simulation of the coolant flow at different tilt angles, ranging from 0 to 45 degrees, and discretized at intervals of 5 degrees. The finite element method was used to solve the Navier-Stokes equations and energy equation partial differential equations containing velocity, pressure and temperature, obtain the velocity field and temperature field distribution of the coolant at each tilt angle, calculate the trajectory line and temperature contour line of the fluid particles, and use the cubic spline interpolation algorithm to smooth the data between discrete points. Shannon entropy is used to calculate the uniformity index of temperature distribution, discretize the temperature field into n intervals, calculate the probability pi of each interval, and the entropy value S = -Σ(pi*log2(pi)). By comparing the entropy values ​​at different tilt angles, the uniformity of temperature distribution is quantitatively evaluated. At the same time, the flow resistance coefficient is calculated to determine the smoothness of coolant flow. The uniformity index and the smoothness index are weighted and summed to obtain a comprehensive evaluation index. When the local temperature exceeds the average temperature by 5°C or the flow rate is lower than the average flow rate by 50%, it is determined to be a heat dissipation risk area. Based on the support vector machine algorithm, the radial basis function kernel is used to optimize the penalty parameter C and the kernel parameter γ through the grid search method. The tilt angle and the comprehensive evaluation index are used as input features to train the heat dissipation risk prediction model, and the optimal parameters are determined through cross-validation. The trained model is used to predict and identify the heat dissipation risk area at the current tilt angle.

[0034] The three-dimensional model constructed by SolidWorks includes 100 21700 lithium battery cells arranged in a 10×10 matrix. The coolant flow channel is 3mm wide, 2mm high, and 2m long. ANSYS Fluent uses the k-ε turbulence model, with about 2 million grid cells, a solution time step of 0.01s, and a total simulation time of 300s. The Reynolds number Re=5000 and the Prandtl number Pr=7 in the Navier-Stokes equations. The cubic spline interpolation uses natural boundary conditions with 100 nodes. In the Shannon entropy calculation, the temperature field is discretized into 20 equally spaced intervals with an interval width of 1℃. The flow resistance coefficient is calculated by dividing the total pressure drop by the dynamic pressure. Among the comprehensive evaluation indicators, the weight of temperature uniformity is 0.6 and the weight of flow smoothness is 0.4. The support vector machine uses a radial basis function kernel, the search range of the penalty parameter C is [2^-5, 2^15], the search range of the kernel parameter γ is [2^-15, 2^3], and the step size is a power of 2. The grid search uses 5-fold cross validation. The final model has an accuracy of 92% on the test set. The simulation results show that at an inclination angle of 30 degrees, the temperature distribution is the most uniform (entropy value 0.95) and the flow is the smoothest (resistance coefficient 0.35). The heat dissipation risk area is mainly concentrated at the bottom of the battery pack near the outlet, accounting for 8% of the total area. The model can predict the heat dissipation risk at different inclination angles in real time, providing data support for cooling system optimization.

[0035] Step S106, for the heat dissipation risk area, the speed and flow of the coolant pump are controlled to enhance local heat dissipation, and / or the speed of the electronic fan on the battery pack shell is controlled to enhance air convection, and / or the active thermal management system is controlled to balance the temperature distribution inside the battery pack, and when the heat dissipation risk is greater than a preset safety threshold, the charge and discharge power of the battery cells in the heat dissipation risk area is temporarily reduced.

[0036] The temperature distribution data inside the battery pack is obtained, wherein the temperature distribution data includes the temperature data of the heat dissipation risk area; the temperature deviation and the temperature change rate are calculated according to the temperature distribution data, the target speed of the coolant pump is determined by a fuzzy control algorithm, and the output frequency of the coolant pump is adjusted by a frequency converter; the position information of the heat dissipation risk area is obtained, the fan speed increment is calculated by a PID controller, and a PWM signal is generated according to the fan speed increment, wherein the PWM signal is used to control the fan speed; if the temperature of the heat dissipation risk area exceeds a preset threshold, the active thermal management system is started, and the active thermal management system includes a thermoelectric cooler and a heating film; it is determined whether the temperature of the heat dissipation risk area exceeds the heat dissipation risk safety threshold, and if it exceeds the heat dissipation risk safety threshold, the charge and discharge power of the battery cells in the heat dissipation risk area is controlled by the battery management system.

[0037] Specifically, a temperature sensor array is used to monitor the temperature distribution inside the battery pack in real time, and the temperature data of the heat dissipation risk area is obtained through the data acquisition module. The target speed of the coolant pump is calculated based on the temperature deviation and change rate using the fuzzy control algorithm. Five temperature deviation fuzzy sets and five temperature change rate fuzzy sets are set. The Mamdani reasoning method is used to adjust the output frequency of the coolant pump through the inverter. The frequency adjustment range is 20Hz to 60Hz to achieve dynamic adjustment of the coolant flow rate. According to the location information of the heat dissipation risk area, the speed of the electronic fan at the corresponding position on the battery pack shell is controlled. The proportional integral differential (PID) controller is used. The PID parameters are determined using the Ziegler-Nichols tuning method. The difference between the target temperature and the actual temperature is used as input to calculate the speed increment of the fan. The fan speed is controlled by a pulse width modulation (PWM) signal. The PWM frequency is 25kHz and the duty cycle range is 20% to 100%, which enhances local air convection. The active thermal management system is started. The internal temperature of the battery pack is locally adjusted through the combination of thermoelectric coolers and heating films. The temperature distribution surface is fitted using the least squares method, and quadratic polynomial fitting is adopted. The normal equation is solved using QR decomposition, the temperature gradient is calculated, and the gradient descent algorithm is used to optimize the power distribution of the thermoelectric cooler and the heating film to achieve the balance of the internal temperature of the battery pack and evaluate the temperature balance effect. A heat dissipation risk safety threshold is set. When the heat dissipation risk exceeds the preset threshold, the battery management system (BMS) is used to control the charge and discharge power of the battery cells in the heat dissipation risk area. The adaptive threshold algorithm is used to dynamically adjust the power limit. The exponentially weighted moving average method is used to calculate the dynamic threshold with a weight factor of 0.8. The future temperature is predicted according to the temperature change trend, and the power is adjusted in advance to prevent the temperature from continuing to rise. Coolant flow regulation is preferred, followed by fan speed control, then active thermal management system, and finally power limitation. The temperature sensor array consists of 100 PT100 platinum resistance sensors with an accuracy of

[0038] ±0.1℃, sampling frequency is 10Hz. The fuzzy control algorithm uses a 5×5 rule matrix, the input variables are temperature deviation -5℃ to +5℃ and temperature change rate -1℃ / s to +1℃ / s, and the output is the coolant pump speed increment -300rpm to +300rpm. The inverter maps the 0-10V control signal to an output frequency of 20-60Hz. The initial parameters of the PID controller are set to Kp=10, Ki=0.5, Kd=2 by the Ziegler-Nichols method, and the output is updated every 100ms. The fan PWM control uses an STM32F103 microcontroller to generate a 25kHz PWM signal with a duty cycle accuracy of 0.1%. The temperature distribution surface fitting uses a quadratic polynomial with 9 coefficients, and the normal equation is solved using QR decomposition, with a calculation time of less than 10ms. The learning rate of the gradient descent algorithm is set to 0.01, and the maximum number of iterations is 100. The cooling power range of the thermoelectric cooler is 0-50W, and the heating power range of the heating film is 0-30W. The adaptive threshold algorithm uses 30 minutes of temperature data to calculate the dynamic threshold, and the initial threshold is set to 45°C. BMS adjusts the charge and discharge power according to the dynamic threshold, with a maximum limit of 50% of the rated power. It is executed once every 1 second, with the priority order being coolant flow regulation, fan speed control, active thermal management system, and power limitation, with weights of 0.3, 0.2, 0.1, and 0.4, respectively. In the simulation test, the maximum temperature difference of a 100Ah battery pack was reduced from 8°C to 3°C, and the heat dissipation efficiency was improved by 30%.

[0039] Step S107, a battery life prediction model is established based on the battery power, temperature, charging process control and cooling system optimization data. The health status and remaining life of the battery are evaluated through machine learning algorithms, and the location and severity of thermal imbalance faults are included in the evaluation indicators to formulate battery pack maintenance and replacement strategies to ensure the safe and reliable operation of the electric motorcycle.

[0040] The battery power, temperature distribution, charging process control parameters and cooling system optimization data generated by the battery management system are obtained, and the data is cleaned by the preprocessing module. Based on the cleaned data, a random forest algorithm is used to construct a battery life prediction model, and the prediction model divides the data into a training set and a test set. For the prediction model, the thermal imbalance index is calculated, and the thermal imbalance index is equal to the sum of the temperature deviation multiplied by the position weight divided by the total number of batteries, and the thermal imbalance index is input into the prediction model as a new feature. The support vector regression model is used to predict the battery capacity attenuation trend. The prediction results of the support vector regression model are combined with the results of the battery life prediction model to obtain a comprehensive evaluation result of the battery health status. According to the comprehensive evaluation results, a maintenance and replacement strategy for the battery pack is formulated. The strategy divides the battery pack into three levels: normal, maintenance required, and replacement required. If the battery pack health status is lower than the preset threshold, it is determined to require maintenance or replacement.

[0041] Specifically, the battery power, temperature distribution, charging process control parameters and cooling system optimization data are obtained from the battery management system. The raw data is cleaned through the data preprocessing module, and moving average filtering is used for denoising, Z-score standardization, principal component analysis is used for feature extraction, Fourier transform is used to extract temperature fluctuation features, and the sliding window method is used to calculate the charge and discharge rate to generate a data set containing battery cycle times, average discharge depth, maximum temperature, temperature fluctuation amplitude, charging rate and cooling efficiency. The random forest algorithm is used to build a battery life prediction model, and the number of decision trees is set to 100, the maximum depth is 10, and the minimum number of leaf node samples is 5. The processed data set is divided into a training set and a test set. A 5-fold cross validation is used, and the grid search method is used to optimize the random forest parameters. The key factors affecting the battery life are determined through feature importance analysis, and the preliminary evaluation results of the battery health status and remaining life are obtained, and the interface of the thermal imbalance feature is reserved. According to the location and severity of the thermal imbalance fault, the thermal imbalance index calculation formula is designed. The thermal imbalance index is equal to the sum of the temperature deviation multiplied by the position weight divided by the total number of batteries. The thermal imbalance index is input into the prediction model as a new feature, and the model is retrained and its impact on the battery life prediction is evaluated. The evaluation results of the battery health status and remaining life are updated. The support vector regression model is used to predict the battery capacity decay trend, and the battery health status is comprehensively evaluated in combination with the prediction results of the random forest. Based on the updated evaluation results, the maintenance and replacement strategy of the battery pack is formulated. By setting the health status threshold and the remaining life threshold, based on the statistical analysis of historical data and the battery capacity decay curve, the health status threshold is set to 85% and the remaining life threshold is set to 200 cycles. The battery pack is divided into three levels: normal, maintenance required, and replacement required. For the battery pack that needs maintenance, specific maintenance suggestions are generated, including balanced charging, cooling system cleaning and other operations. For the battery pack that needs to be replaced, the optimal replacement time is calculated and a replacement plan is generated. The battery management system collects data every 5 minutes, including the voltage, current, temperature and charge and discharge status of 100 battery cells. The moving average filter uses a 10-point window, and the Z-score standardization normalizes all features to a distribution with a mean of 0 and a standard deviation of 1. The principal component analysis retains 95% of the variance information, reducing the original 30 features to 12. The Fourier transform uses a 1024-point FFT to extract the first 10 harmonic amplitudes as temperature fluctuation features. The random forest model training uses 80% of the data and the test set is 20%. Grid search optimization parameter range: the number of trees

[0042] [50,100,150], maximum depth [5,10,15], minimum number of leaf node samples [3,5,7]. Feature importance analysis shows that the number of cycles, maximum temperature and average discharge depth are the top three influencing factors. In the calculation of thermal imbalance index, the position weight is set according to the distance from the battery center.

[0043] [0.5, 0.7, 1.0, 1.3, 1.5]. The support vector regression model uses the RBF kernel function and optimizes the parameters C and gamma through cross validation. The health status assessment comprehensively considers the capacity decay rate and the predicted number of remaining cycles, with weights of 0.6 and 0.4 respectively. The maintenance strategy is set according to the health status: 85%-90% for balanced charging, 80%-85% for cleaning the cooling system, and replacement of batteries below 80%. The optimal replacement time is based on the trade-off between battery cost and performance loss, calculated using a dynamic programming algorithm. Updates to the battery pack status report and maintenance plan are performed once a day. In practical applications, this method extends the service life of a 100Ah battery pack by 20% and reduces maintenance costs by 15%.

[0044] The above embodiment is only one of the preferred implementation modes of the present invention and should not be used to limit the protection scope of the present invention. Any changes or modifications that are made to the main design concept and spirit of the present invention and have no substantive significance, and the technical problems they solve are still consistent with the present invention, should be included in the protection scope of the present invention.

Claims

1. A charging control method for an electric motorcycle, characterized in that: The method comprises: When the electric motorcycle is charging, the tilt angle data of the side stand of the electric motorcycle is obtained. If the tilt angle is within a preset angle range, it is determined that the vehicle body is in a tilted state when charging, and a monitoring program for the flow state of the battery pack coolant is triggered; The flow sensor installed inside the battery pack collects the flow rate data of the coolant in different areas in real time. Based on the preset flow rate threshold, it determines whether there is an area with uneven coolant flow and identifies high-risk areas with uneven coolant distribution. At the same time, the battery management system obtains the power data of each battery cell in the battery pack. The temperature sensor array collects the temperature data of the battery cells in the area with uneven coolant distribution in real time, performs gradient analysis on the temperature data, determines the location and severity of the thermal imbalance fault, and provides data support for subsequent charging strategy adjustment and cooling system optimization. According to the location and severity of the thermal imbalance fault, the charging strategy and cooling system operating parameters are dynamically adjusted. For battery cells with severe thermal imbalance faults and low power, the strategy of reducing charging current and enhancing local cooling is adopted. For battery cells with severe thermal imbalance faults and high power, the strategy of extending charging time and enhancing local cooling is adopted. Simulate the temperature distribution and flow trajectory of the coolant at different tilt angles, and determine the heat dissipation risk area at the current tilt angle by comparing and analyzing the uniformity of temperature distribution and smoothness of flow; For heat dissipation risk areas, control the speed and flow of the coolant pump to enhance local heat dissipation, and / or control the speed of the electronic fan on the battery pack housing to enhance air convection, and / or control the active thermal management system to balance the temperature distribution inside the battery pack, and temporarily reduce the charge and discharge power of the battery cells in the heat dissipation risk area when the heat dissipation risk is greater than the preset safety threshold; A battery life prediction model is established based on data from battery charge, temperature, charging process control, and cooling system optimization. The battery's health status and remaining life are evaluated through a machine learning algorithm. The location and severity of thermal imbalance failures are included in the evaluation indicators, and a battery pack maintenance and replacement strategy is formulated to ensure the safe and reliable operation of electric motorcycles.

2. The method according to claim 1, characterized in that When the electric motorcycle is charging, the tilt angle data of the side bracket of the electric motorcycle is obtained. If the tilt angle is within a preset angle range, it is determined that the vehicle body is in a tilted state when charging, and a monitoring program for the flow state of the battery pack coolant is triggered, including: Obtaining raw data collected by a three-axis acceleration sensor installed on a main control panel of the vehicle body, calculating the inclination angle value of the side bracket of the electric motorcycle using the cosine theorem, and transmitting the inclination angle value to a vehicle control unit; According to the tilt angle value received by the vehicle control unit, the tilt angle value is compared with a preset angle range, and if the tilt angle value falls within the preset angle range, it is determined that the vehicle body is in a tilt state in a charging state; After determining that the vehicle body is in a tilted state while charging, the battery pack coolant flow state monitoring program is started, and real-time flow data is collected through the flow sensor installed on the coolant pipeline, combined with the battery temperature information obtained by the temperature sensor installed on the battery pack shell; The coolant flow rate data and battery temperature data are obtained, and a cooling effect evaluation model is established using a support vector regression algorithm. The coolant flow rate and battery temperature are used as input features to obtain a cooling effect score. According to the cooling effect score output by the cooling effect evaluation model, the rotation speed of the coolant pump is controlled by a pulse width modulation signal to achieve precise regulation of the coolant flow rate to optimize the battery temperature control during the charging process.

3. The method according to claim 1, characterized in that The flow sensor installed inside the battery pack collects the flow rate data of the coolant in different areas in real time, determines whether there is an area with uneven coolant flow according to a preset flow rate threshold, and identifies high-risk areas with uneven coolant distribution, and obtains the power data of each battery cell in the battery pack through the battery management system, including: Obtain coolant flow rate data collected by multiple thermal flow sensors inside the battery pack; Comparing the flow rate data with a preset flow rate threshold, if the flow rate in a certain area is lower than the threshold, it is determined that the coolant flow in the area is uneven; Clustering algorithm is used to group the areas with uneven flow and obtain high-risk areas with uneven coolant distribution; Acquire power data of each battery cell in the battery pack from the battery management system through the communication interface, perform correlation analysis on the power data and the coolant flow rate data, and use a decision tree algorithm to establish a relationship model between the power of the battery cell and the cooling demand; According to the cooling demand prediction value output by the relationship model, the increase or decrease ratio of the coolant flow required for each area is calculated, and the ball valve in the battery pack cooling system is adjusted to increase the coolant flow in high-risk areas while reducing the flow in other areas to achieve dynamic balanced distribution of coolant inside the battery pack.

4. The method according to claim 1, characterized in that The temperature sensor array is used to collect the temperature data of the battery cells in the area where the coolant is unevenly distributed in real time, and the temperature data is gradient analyzed to determine the location and severity of the thermal imbalance fault, so as to provide data support for the subsequent charging strategy adjustment and cooling system optimization, including: Obtain temperature data of the area where the coolant is unevenly distributed; Preprocessing the temperature data, including smoothing the temperature data using a sliding window method; Calculate the mean and standard deviation of the temperature data based on the window size and step size; Eliminate abnormal data points to obtain a smoothed temperature data series; Determine the thermal imbalance fault area, including: using the finite difference method to calculate the temperature gradient; If the temperature gradient exceeds a preset threshold, it is determined to be a thermal imbalance fault area; Determine the severity of the fault based on the magnitude of the temperature gradient; Identify thermal imbalance failure modes, including: deep learning of temperature data using convolutional neural networks; Supervised learning through historical failure data; Combine the output of temperature gradient analysis to identify thermal imbalance failure modes of hot spots, hot zones, and local overheating; Generate a thermal imbalance fault report, including: determining the location and severity of the thermal imbalance fault based on the thermal imbalance fault mode and fault severity; Adjust charging current and coolant flow distribution based on thermal imbalance fault reporting.

5. The method according to claim 1, characterized in that The method of dynamically adjusting the charging strategy and the operating parameters of the cooling system according to the location and severity information of the thermal imbalance fault, and taking the strategy of reducing the charging current and enhancing the local cooling for the battery cells with severe thermal imbalance fault and low power, and taking the strategy of extending the charging time and enhancing the local cooling for the battery cells with severe thermal imbalance fault and high power, includes: Obtain the location and severity of thermal imbalance faults of battery cells, and read the real-time power data of battery cells from the battery management system; Based on the thermal imbalance fault information and power data, a filtering algorithm is used to fuse the data to obtain a comprehensive data report containing the battery ID, fault location coordinates, severity level, and current power percentage; Based on the comprehensive data report, the charging strategy is dynamically adjusted through a fuzzy logic controller, wherein the input of the controller is the fault severity and the power level, and the output of the controller is the charging current adjustment value and the charging time adjustment coefficient; If the battery cell thermal imbalance fault is serious and the battery power is low, the charging current is reduced, and the charging current adjustment range is from the preset minimum value to the maximum value; For battery cells with severe thermal imbalance faults and high power, the optimal charging time is calculated, and the charging time is extended from the preset minimum multiple to the maximum multiple of the original planned charging time; After receiving the local cooling enhancement command, the fuzzy adaptive PID controller is called to dynamically adjust the coolant flow and temperature according to the location and severity of the thermal imbalance fault. The PID parameters are adjusted in real time according to the temperature deviation and change rate.

6. The method according to claim 1, characterized in that The simulation of the temperature distribution and flow trajectory of the coolant at different tilt angles, and determining the heat dissipation risk area at the current tilt angle by comparing and analyzing the uniformity of the temperature distribution and the smoothness of the flow, includes: Obtaining actual size parameters of the battery pack, flow channel, and heat sink, and constructing a three-dimensional geometric model based on the parameters; Setting a coolant flow channel, a battery cell position, and a radiator position on the three-dimensional geometric model to obtain a complete battery cooling system model; ANSYS Fluent software is used to perform numerical simulation on the battery cooling system model at different tilt angles, and the Navier-Stokes equation and energy equation partial differential equation system including velocity, pressure and temperature are solved to obtain the velocity field and temperature field distribution of the coolant; According to the temperature field distribution, the uniformity index of the temperature distribution is calculated using Shannon entropy, wherein the uniformity index is obtained by discretizing the temperature field into n intervals, calculating the probability pi of each interval, and obtaining an entropy value; Calculate the resistance coefficient of coolant flow, judge the smoothness of coolant flow, and perform weighted summation of uniformity index and smoothness index to obtain a comprehensive evaluation index; Based on the support vector machine algorithm, the tilt angle and the comprehensive evaluation index are used as input features to train a heat dissipation risk prediction model. The model optimizes the penalty parameter C and the kernel parameter γ through the radial basis function kernel and the grid search method to predict and identify the heat dissipation risk area under the current tilt angle.

7. The method according to claim 1, characterized in that The above mentioned control measures for the heat dissipation risk area include controlling the speed and flow of the coolant pump to enhance local heat dissipation, and / or controlling the speed of the electronic fan on the battery pack housing to enhance air convection, and / or controlling the active thermal management system to balance the temperature distribution inside the battery pack, and temporarily reducing the charge and discharge power of the battery cells in the heat dissipation risk area when the heat dissipation risk is greater than a preset safety threshold, including: Acquire temperature distribution data inside the battery pack, wherein the temperature distribution data includes temperature data of a heat dissipation risk area; Calculate the temperature deviation and the temperature change rate according to the temperature distribution data, determine the target speed of the coolant pump by using a fuzzy control algorithm, and adjust the output frequency of the coolant pump by using a frequency converter; Acquire location information of the heat dissipation risk area, use a PID controller to calculate the fan speed increment, and generate a PWM signal according to the fan speed increment, wherein the PWM signal is used to control the fan speed; If the temperature of the heat dissipation risk area exceeds a preset threshold, an active thermal management system is activated, wherein the active thermal management system includes a thermoelectric cooler and a heating film; It is determined whether the temperature of the heat dissipation risk area exceeds the heat dissipation risk safety threshold. If it exceeds the heat dissipation risk safety threshold, the battery management system is used to control the charge and discharge power of the battery cells in the heat dissipation risk area.

8. The method according to claim 1, characterized in that The battery life prediction model is established based on the data of battery power, temperature, charging process control and cooling system optimization. The health status and remaining life of the battery are evaluated through machine learning algorithms. The location and severity of thermal imbalance failure are included in the evaluation indicators, and the maintenance and replacement strategy of the battery pack is formulated to ensure the safe and reliable operation of the electric motorcycle, including: Acquire battery power, temperature distribution, charging process control parameters and cooling system optimization data generated by the battery management system, and clean the data through a preprocessing module; Based on the cleaned data, a battery life prediction model is constructed using a random forest algorithm, wherein the prediction model divides the data into a training set and a test set; For the prediction model, calculating a thermal imbalance index, where the thermal imbalance index is equal to the sum of the temperature deviation multiplied by the position weight divided by the total number of batteries, and inputting the thermal imbalance index into the prediction model as a new feature; A support vector regression model is used to predict the battery capacity attenuation trend, and the prediction result of the support vector regression model is combined with the result of the battery life prediction model to obtain a comprehensive evaluation result of the battery health status; Based on the comprehensive evaluation results, a maintenance and replacement strategy for the battery pack is formulated. The strategy divides the battery pack into three levels: normal, maintenance required, and replacement required. If the health status of the battery pack is lower than a preset threshold, it is determined to require maintenance or replacement.

Images