Portable motorcycle emergency charging scheme generation method based on plug-in design

Abstract

The invention provides a plug-in type design-based portable motorcycle emergency charging scheme generation method, which comprises the following steps of: acquiring a capacitor voltage parameter and a capacitor internal resistance parameter of each capacitor layer, and training a nonlinear mapping model by taking temperature and current data in a charging and discharging process as input and taking the voltage and internal resistance parameter of each capacitor layer as output; through cross validation, optimizing capacitance, temperature and current parameters of the nonlinear mapping model; in the charging and discharging process of the plug-in type charging module, temperature and current data of each capacitor layer are collected in real time and input into the nonlinear mapping model, and voltage and internal resistance parameters of each capacitor layer are predicted to obtain dynamic changes of the parameters of the capacitor layers in the charging and discharging process.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a method for generating an emergency charging solution for a portable motorcycle based on a plug-in design. Background Art

[0002] When a motorcycle is running low on power during long-distance outdoor driving, emergency charging is required. If the voltage and internal resistance of each capacitor layer in the motorcycle charging module are unbalanced, the charging efficiency will be too slow and the purpose of emergency charging cannot be achieved. Therefore, in the generation of an emergency charging scheme based on a plug-in portable motorcycle, it is necessary to balance the voltage, internal resistance and other parameters in the layered capacitor structure of the plug-in charging module, and the balanced control faces many technical challenges. First, the voltage and internal resistance parameters of different capacitor layers are affected by temperature and current during the charging and discharging process, showing complex nonlinear characteristics. This nonlinear characteristic makes it difficult to accurately control the charging and discharging state of each capacitor layer, thereby affecting the performance and reliability of the entire charging module. Secondly, in order to achieve consistent discharge depth of each layer of capacitor, it is necessary to monitor and dynamically balance the voltage and internal resistance parameters of each layer of capacitor in real time. However, in the actual working environment of the charging module, temperature changes and fluctuations in charge and discharge currents will interfere with the accuracy of capacitor parameter measurement, increasing the difficulty of designing a balanced control strategy. In addition, in the plug-in design, in order to adapt to the use scenario of frequent plugging and unplugging, higher requirements are placed on the integrated design of the charging module. In summary, how to achieve accurate modeling, parameter identification and balanced control of the layered capacitor structure under complex working conditions is a key issue that needs to be overcome in the portable motorcycle emergency charging technology. Summary of the invention

[0003] The present invention provides a method for generating an emergency charging scheme for a portable motorcycle based on a plug-in design, which mainly includes:

[0004] Obtaining the capacitance voltage parameters and capacitance internal resistance parameters of each capacitance layer, taking the temperature and current data in the charging and discharging process as input, and taking the voltage and internal resistance parameters of each capacitance layer as output, training the nonlinear mapping model, and optimizing the capacitance, temperature and current parameters of the nonlinear mapping model through cross-validation;

[0005] During the charging and discharging process of the plug-in charging module, the temperature and current data of each capacitor layer are collected in real time and input into the nonlinear mapping model to predict the voltage and internal resistance parameters of each capacitor layer to obtain the dynamic changes of the capacitor layer parameters during the charging and discharging process;

[0006] The predicted voltage of each capacitor layer is compared with the preset voltage balance threshold to determine whether the voltage of each capacitor layer is balanced. If not, the voltage of each layer is adjusted through the energy transfer circuit until a balanced state is reached. The difference between the internal resistance of each capacitor layer and the predicted internal resistance parameter is calculated, and the internal resistance compensation value is determined according to the difference. The capacitor layer whose internal resistance difference is greater than the preset internal resistance difference is compensated;

[0007] The calculated voltage difference and internal resistance difference of each capacitor layer are used as the input of the preset fuzzy control algorithm, and the charge and discharge current adjustment amount of each capacitor layer is used as the output. By fuzzifying the input variables, the target charge and discharge current control amount of each capacitor layer is output, and the target charge and discharge current control amount is transmitted to the current regulation circuit of the plug-in charging module to adjust the charge and discharge current of each capacitor layer in real time;

[0008] During the emergency charging process of a plug-in portable motorcycle, the temperature, current, voltage and internal resistance parameters of each capacitor layer of the plug-in charging module are continuously monitored. According to the prediction results of the nonlinear mapping model and the target charge and discharge current control amount, the charge and discharge current of each capacitor layer is dynamically optimized to achieve balanced charge and discharge of the layered capacitor structure.

[0009] During the emergency charging process based on the plug-in portable motorcycle, the motorcycle power is also monitored in real time, and an emergency charging plan for the portable motorcycle is generated based on the motorcycle power. If the motorcycle power is monitored to be in a preset low power range, the balancing control of the discharge of the layered capacitor structure of the plug-in charging module is triggered. If the monitored motorcycle power is higher than the preset high power threshold, the frequency of balancing control is gradually reduced until the motorcycle power reaches full charge, and the balancing control is stopped and charging is terminated.

[0010] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0011] The present invention discloses a method for generating an emergency charging scheme for a portable motorcycle based on a plug-in design. The method predicts the voltage and internal resistance parameters of each capacitor layer in real time by establishing a nonlinear mapping model between the capacitor layer parameters and the temperature and current, and dynamically adjusts the charge and discharge current according to the prediction results. During the charging process, the present invention continuously monitors the power of the motorcycle and adaptively adjusts the balancing control strategy according to the power status. When the power is in the low power range, the balanced discharge control of the layered capacitor structure is triggered; when the power exceeds the high power threshold, the balancing control frequency is gradually reduced until it is fully charged and stopped. The present invention realizes precise charge and discharge regulation of each capacitor layer through a fuzzy control algorithm, effectively solves the problems of voltage imbalance and internal resistance difference in the plug-in portable charging process, and improves the charging efficiency and battery life. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1The present invention is a flow chart of a method for generating an emergency charging solution for a portable motorcycle based on a plug-in design.

[0013] Figure 2 It is a schematic diagram of a method for generating an emergency charging solution for a portable motorcycle based on a plug-in design according to the present invention.

[0014] Figure 3 It is another schematic diagram of a method for generating an emergency charging solution for a portable motorcycle based on a plug-in design according to the present invention. DETAILED DESCRIPTION

[0015] In order to further understand the content of the present invention, the present invention is described in detail in conjunction with the accompanying drawings and embodiments. The present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are only used to explain the relevant inventions, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the invention are shown in the accompanying drawings.

[0016] like Figure 1-3 In this embodiment, a method for generating an emergency charging solution for a portable motorcycle based on a plug-in design may specifically include:

[0017] Step S101, obtain the capacitor voltage parameters and capacitor internal resistance parameters of each capacitor layer, use the temperature and current data during the charging and discharging process as input, and use the voltage and internal resistance parameters of each capacitor layer as output to train a nonlinear mapping model, and optimize the capacitance, temperature and current parameters of the nonlinear mapping model through cross-validation.

[0018] The temperature signal and the current signal during the charging and discharging process are acquired from a parameter acquisition device, and the parameter acquisition device records the voltage value and the internal resistance value at the corresponding moment; according to the temperature signal and the current signal, the data cleaning unit removes abnormal fluctuation points to obtain a standardized training data set; for the standardized training data set, a data partitioning unit is used to perform cross-grouping, and a training data subset and a verification data subset are generated according to the proportion of the training set; a nonlinear mapping function from the temperature signal and the current signal to the voltage value and the internal resistance value is established through a neural network mapper, and the parameters of the nonlinear mapping function are trained using the training data subset to obtain a first round of optimization parameter set; if the average error value on the verification data subset is higher than the average error threshold, the gradient descent optimizer is used to iteratively update the first round of optimization parameter set until the average error value is lower than the average error threshold.

[0019] Specifically, the temperature and current signals during the charging and discharging process are obtained from the parameter acquisition device, and the voltage and internal resistance values ​​at the corresponding time are recorded using the same acquisition device. The abnormal fluctuation points are removed by the data cleaning unit to generate a standardized training data set. The standardized training data set is cross-grouped by a data partitioning unit, and a training data subset and a verification data subset are generated according to the partitioning rule that the training set accounts for 80%, and a nonlinear mapping function from temperature, current to voltage and internal resistance is established by a neural network mapper. For the nonlinear mapping function, parameter training is performed using the training data subset, and the mapping error value on the verification data subset is calculated by the verification error calculator. The capacitance coefficient, temperature coefficient, and current coefficient in the mapping function are optimized and adjusted according to the grid search method to obtain the first round of optimized parameter set. Based on the first round of optimized parameter set, the parameters are iteratively updated in the gradient descent optimizer, and the remaining unused verification data is used for cross-validation until the average error value on all verification data subsets is lower than the preset threshold, and the final optimized nonlinear mapping parameter set is output. The parameter acquisition device synchronously collects the key parameters of the capacitor layer during the charging and discharging process through a high-precision sampling circuit. The temperature signal sampling frequency is set to 10 Hz, the measurement accuracy is 0.1 degrees Celsius, and the measurement range covers -40 to 125 degrees Celsius. The current signal sampling frequency is 1 kHz, the measurement accuracy is 0.01 amperes, and the measurement range is 0 to 10 amperes. The voltage signal sampling accuracy is 0.01 volts, and the internal resistance measurement adopts the AC impedance method with a measurement frequency of 1 kHz. The data cleaning unit processes the collected raw data for outliers. When the temperature change rate exceeds 5 degrees Celsius per second or the current mutation exceeds 2 amperes per microsecond, the data point is marked as an outlier and removed. The five-fold cross-grouping method divides the standardized data set into 5 equal parts, selects 4 parts as training sets each time, and the remaining 1 part as a validation set to achieve 5 rounds of cross-validation. In actual operation, for a total of 10,000 sets of valid data, each round of training set contains 8,000 sets of data and the validation set contains 2,000 sets of data. The neural network mapper adopts a three-layer structure. The input layer contains two nodes, temperature and current, the hidden layer uses 16 nodes, the output layer contains two nodes, voltage and internal resistance, and the activation function uses the hyperbolic tangent function. During the parameter training process of the nonlinear mapping function, the initial value of the capacitance coefficient is set to 0.1, the initial value of the temperature coefficient is set to 0.01, and the initial value of the current coefficient is set to 0.05. The grid search method sets the search range of the capacitance coefficient to 0.05 to 0.5 with a step size of 0.05; the search range of the temperature coefficient is 0.005 to 0.05 with a step size of 0.005; the search range of the current coefficient is 0.02 to 0.2 with a step size of 0.02. By calculating the root mean square error on the validation set, the parameter combination that minimizes the error is found. Based on the first round of optimization, the gradient descent optimizer adopts an adaptive learning rate, and the initial learning rate is set to 0.01. When the error does not decrease in three consecutive rounds of validation, the learning rate is reduced to 0.5 times the original.During the verification process, the relative error of voltage prediction is required to be less than 3%, and the relative error of internal resistance prediction is required to be less than 5%. In the actual verification, when the temperature is 25 degrees Celsius and the charging current is 1 ampere, the voltage prediction value is 2.7 volts, the measured value is 2.75 volts, and the relative error is 1.8%; the internal resistance prediction value is 150 milliohms, the measured value is 155 milliohms, and the relative error is 3.2%. Under extreme working conditions, the internal resistance prediction error at -20 degrees Celsius increases slightly, but still remains within 5%, meeting the requirements of engineering applications. The nonlinear mapping model optimized by cross-validation shows good prediction ability under different temperature and current conditions, especially in the common operating temperature range of capacitors, 0 to 60 degrees Celsius, with higher prediction accuracy. The nonlinear mapping model can accurately reflect the influence of temperature and current on the performance of the capacitor layer. The model prediction results can be used to optimize the working parameter settings of the capacitor and select the appropriate charge and discharge current at different ambient temperatures, which not only ensures the charge and discharge efficiency, but also avoids overheating of the capacitor due to excessive charge and discharge current. The model can effectively reflect the influence of temperature on internal resistance and the nonlinear characteristics of large current charging and discharging. In practical applications, this model can be used to estimate capacitor parameters online and determine capacitor performance status.

[0020] Step S102, during the charging and discharging process of the plug-in charging module, the temperature and current data of each capacitor layer are collected in real time, input into a nonlinear mapping model, and the voltage and internal resistance parameters of each capacitor layer are predicted to obtain the dynamic changes of the capacitor layer parameters during the charging and discharging process.

[0021] Receive the surface temperature distribution data and current change data of the capacitor layer collected by the plug-in charging module, and generate a real-time monitoring data set with a timestamp according to the fixed sampling time interval. Use a data preprocessing unit to standardize the real-time monitoring data set, and predict the voltage parameters and internal resistance parameters of the capacitor layer through a nonlinear mapping model based on the standardized data. Extract timing features based on the predicted voltage parameters and internal resistance parameters, and a state identifier generates a timing correspondence between the voltage parameters, internal resistance parameters and the temperature data and current data based on the timing features. Use a Kalman filter to smooth the voltage parameters and internal resistance parameters in the timing correspondence, and obtain a dynamic change curve of the capacitor layer parameters based on the smoothed parameters combined with the working state of the plug-in charging module.

[0022] Specifically, the temperature distribution data of the surface of the capacitor layer is collected in real time by the multi-channel measuring device of the plug-in charging module, and the current change data of the charging and discharging process is recorded by a high-precision current sensor, and a real-time monitoring data set with a time stamp is generated according to a fixed sampling time interval. For the real-time monitoring data set, the temperature distribution data and the current data are standardized by a data preprocessing unit, and input into a nonlinear mapping model, and the voltage and internal resistance parameters of the capacitor layer are continuously predicted according to a preset sampling period. The voltage timing characteristics and the internal resistance timing characteristics are extracted from the continuous prediction results, and the charging and discharging state of the capacitor layer is judged in real time by a state identifier, and the timing correspondence between the voltage, internal resistance parameters and the temperature and current data is generated by a dynamic feature extractor. Based on the timing correspondence, the voltage and internal resistance prediction sequences are smoothed in the Kalman filter, and the prediction parameters are corrected in combination with the working state of the plug-in charging module to obtain the dynamic change curve of the capacitor layer parameters. The multi-channel measuring device of the plug-in charging module arranges 8 temperature sensing points on the surface of the capacitor layer, and adopts a platinum resistance temperature sensor with a sampling frequency of 10 Hz and a measurement accuracy of 0.1 degrees Celsius. The high-precision current sensor uses a Hall current sensor with a sampling frequency of 1 kHz and a measurement accuracy of 0.01 amperes. The timestamp accuracy in the real-time monitoring data set is at the millisecond level, recording the specific collection time of each sampling point. The data preprocessing unit normalizes the temperature data, mapping the temperature range of -40 to 125 degrees Celsius to the interval of 0 to 1, and the current data maps the range of -10 to 10 amperes to the interval of -1 to 1. The nonlinear mapping model uses a preset sampling period of 100 milliseconds to complete a prediction calculation of the voltage and internal resistance parameters once in each sampling period. When the temperature distribution of the capacitor layer is uneven and the maximum temperature difference exceeds 20 degrees Celsius, the model will assign different weights to each temperature measurement point. The voltage time series characteristics include three dimensions: absolute value of voltage, voltage change rate, and voltage fluctuation amplitude, and the internal resistance time series characteristics include three dimensions: absolute value of internal resistance, temperature coefficient of internal resistance, and current coefficient of internal resistance. The state identifier determines the charge and discharge state by the product of voltage and current. Positive values ​​indicate charging state, and negative values ​​indicate discharging state. The dynamic feature extractor uses a 200 millisecond time window to perform sliding average processing on each feature. The state vector of the Kalman filter contains two variables: voltage and internal resistance. The parameters of the observation noise covariance matrix are set according to the sensor accuracy, and the process noise covariance matrix is ​​adaptively adjusted according to the charging and discharging conditions. In practical applications, when the capacitor is in a constant temperature environment of 25 degrees Celsius and charged with a constant current of 1 ampere, the error between the predicted voltage value and the measured value is within 20 millivolts, and the error between the predicted internal resistance value and the measured value is within 10 milliohms. The working states of the plug-in charging module are divided into three types: standby, charging, and discharging. In the charging state, the input current is positive and the voltage gradually rises to the rated value of 2.7 volts. In the discharging state, the output current is negative and the voltage gradually decreases, and when it drops to 2.0 volts, it enters the protection state.The dynamic curve shows that at 25 degrees Celsius, during the 1 ampere constant current charging process, it takes 300 seconds for the capacitor layer voltage to rise from 0 to 2.7 volts, and the internal resistance remains at around 150 milliohms. When the ambient temperature rises to 75 degrees Celsius, charging under the same conditions only takes 250 seconds, and the internal resistance drops to 120 milliohms. The real-time monitoring and prediction of dynamic parameter changes show that there is an obvious nonlinear relationship between the voltage and internal resistance parameters of the capacitor layer and temperature and current. Rising temperature will cause the internal resistance to decrease and the charging and discharging efficiency to improve. However, excessively high temperatures will accelerate the aging of the capacitor. When the temperature exceeds 85 degrees Celsius, the internal resistance value fluctuates abnormally. During high current charging and discharging, due to the influence of Joule heat, the temperature distribution of the capacitor layer is uneven, and the temperature difference at each point can reach 15 degrees Celsius, resulting in significant differences in local internal resistance.

[0023] Step S103, compare the predicted voltage of each capacitor layer with the preset voltage balance threshold to determine whether the voltage of each capacitor layer is balanced. If not, adjust the voltage of each layer through the energy transfer circuit until a balanced state is reached; calculate the difference between the internal resistance of each capacitor layer and the predicted internal resistance parameter, determine the internal resistance compensation value according to the difference, and compensate the capacitor layer whose internal resistance difference is greater than the preset internal resistance difference.

[0024] A voltage detector is used to compare and calculate the predicted voltage value of the capacitor layer with the preset voltage balance threshold value to obtain the voltage deviation and the inter-layer voltage difference value; based on the comparison result of the voltage deviation and the preset threshold range, an energy transfer channel is established between adjacent capacitor layers through an energy transfer controller; for the energy transfer channel, the energy transfer time and the energy transfer current are set according to the inter-layer voltage difference value, and the internal resistance difference value of each capacitor layer is calculated by an internal resistance comparator; for the internal resistance difference value, if it exceeds the preset difference value, the compensation parameter generator generates the internal resistance compensation parameter, and the internal resistance compensation process is tracked in real time through the Kalman filter until the internal resistance difference value drops below the preset difference value.

[0025] Specifically, a voltage detector is used to compare and calculate the predicted voltage value of each capacitor layer with the preset voltage balance threshold, and the voltage deviation of each layer and the voltage difference between layers are obtained by the voltage deviation calculator. If the voltage deviation exceeds the preset threshold range, the capacitor layer is marked as an object to be adjusted. For the object to be adjusted, an energy transfer controller is used to establish an energy transfer channel between adjacent capacitor layers, and the energy transfer time and energy transfer current are set according to the voltage difference between layers until the voltage deviation of each layer drops below the preset threshold. Using the predicted internal resistance parameter, the internal resistance difference value of each capacitor layer is calculated by the internal resistance comparator. If the internal resistance difference value exceeds the preset difference value, the internal resistance compensation parameter and the compensation adjustment curve are generated by the compensation parameter generator. Based on the compensation adjustment curve, the compensation adjustment is applied to the capacitor layer according to the compensation parameter in the internal resistance compensation controller, and the internal resistance change during the compensation process is tracked in real time by the Kalman filter until the internal resistance difference value drops below the preset difference value. The voltage detector measures the voltage of multiple capacitor layers. In a typical three-layer capacitor group, the voltage balance threshold is set to 2.7 volts. The actual measured voltages of the upper, middle and lower layers are 2.9 volts, 2.6 volts and 2.5 volts respectively. The voltage deviation calculator calculates the deviations relative to the balance threshold to be 0.2 volts, -0.1 volts and -0.2 volts respectively. The maximum voltage difference between layers is 0.4 volts, which exceeds the preset threshold of 0.2 volts. The energy transfer controller uses a flying capacitor to transfer energy between adjacent layers. The controller sets an energy transfer current of 500 mA based on the maximum inter-layer voltage difference of 0.4 volts. When transferring energy from the upper layer with the highest voltage to the middle layer, each switching cycle is 100 microseconds. The energy transfer time lasts until the inter-layer voltage difference drops below 0.1 volts. The whole process lasts about 200 milliseconds. The energy transfer from the middle layer to the lower layer uses the same current size, and the process lasts 150 milliseconds. The predicted internal resistance parameters of the capacitor layer collected by the internal resistance comparator are 145 milliohms, 150 milliohms, and 165 milliohms, respectively, and the measured internal resistance values ​​are 148 milliohms, 152 milliohms, and 180 milliohms, respectively. The internal resistance difference values ​​are 3 milliohms, 2 milliohms, and 15 milliohms, respectively, and the third layer exceeds the preset 10 milliohm difference. The compensation parameter generator generates a compensation parameter of 20 milliohms for the third layer, and the compensation adjustment curve adopts a piecewise linear method to complete the compensation process within 5 seconds. After receiving the compensation adjustment curve, the internal resistance compensation controller compensates and adjusts the third layer capacitor by changing the amplitude and duration of the charge and discharge current. During the compensation process, the state estimation equation of the Kalman filter contains two variables: internal resistance value and internal resistance change rate, and the measurement noise standard deviation is set to 5 milliohms. In the early stage of compensation, the internal resistance drops rapidly, and it takes 2 seconds to drop from 180 milliohms to 165 milliohms. It is then slowly adjusted to 158 milliohms within 3 seconds. Finally, the internal resistance difference value drops to 8 milliohms, meeting the preset difference requirements. The entire process of voltage balance and internal resistance compensation demonstrates the characteristics of a series capacitor group. Voltage balance is achieved through energy transfer between upper and lower adjacent layers, preventing the voltage of a single capacitor from being too high.Internal resistance compensation reflects the difference in capacitor performance. A large internal resistance difference indicates that the performance of the capacitor has declined. The internal resistance characteristics are adjusted to be close to the normal value through compensation, thereby improving the consistency of the entire group of capacitors. In actual applications, when the ambient temperature rises from 25 degrees Celsius to 75 degrees Celsius, the internal resistance value of the three-layer capacitor decreases by about 20% overall, but the relative difference remains stable, indicating that the compensation effect is not affected by temperature. During the compensation control process, the two key parameters, voltage and internal resistance, influence each other. The capacitor layer with a larger internal resistance changes its voltage faster during the charging and discharging process, which can easily cause voltage imbalance between layers. By coordinating the timing of voltage balance control and internal resistance compensation control, voltage balance is achieved first, and then internal resistance compensation is performed to avoid mutual interference between the two control processes and ensure the stability of the control effect. The control process adopts a closed-loop method to track parameter changes in real time to ensure control accuracy.

[0026] In step S104, the calculated voltage difference and internal resistance difference of each capacitor layer are used as the input of the preset fuzzy control algorithm, and the charge and discharge current regulation amount of each capacitor layer is used as the output. By fuzzifying the input variables, the target charge and discharge current control amount of each capacitor layer is output, and the target charge and discharge current control amount is transmitted to the current regulation circuit of the plug-in charging module to adjust the charge and discharge current of each capacitor layer in real time.

[0027] The voltage difference and internal resistance difference between each capacitor layer are obtained through a voltage internal resistance detection device, and a fuzzy variable processing unit is used to generate a fuzzy membership value for the voltage difference and the internal resistance difference; based on the fuzzy membership value, a fuzzy control rule base of temperature and current compensation is used to perform fuzzy reasoning operations to obtain the fuzzy control quantity of the charge and discharge current of each capacitor layer; for the current fuzzy control quantity, a centroid method defuzzification unit is used to calculate the real-time charge and discharge current regulation quantity of each capacitor layer, and a charge and discharge current instruction with a limit is generated according to the regulation quantity; the charge and discharge current instruction with a limit is used to generate a switch control signal in a pulse width modulator, and the charge and discharge current is adjusted according to the current detection signal of the plug-in charging module.

[0028] Specifically, the voltage difference and internal resistance difference between each capacitor layer are obtained through a voltage internal resistance detection device, a fuzzy variable processing unit is used to divide the voltage difference and the internal resistance difference into language value intervals, and a membership function calculator is used to generate corresponding fuzzy membership values.

[0029]

[0030] , μV(x) represents the membership function of the voltage difference, x represents the actual value of the voltage difference, a represents the lower threshold of the voltage difference, and b represents the upper threshold of the voltage difference.

[0031]

[0032] , μ R (y) represents the membership function of the internal resistance difference, y represents the actual value of the internal resistance difference, c represents the lower limit threshold of the internal resistance difference, and d represents the upper limit threshold of the internal resistance difference.

[0033]

[0034] , F(z) represents the comprehensive membership value, w i represents the weight coefficient of the i-th membership function, μ i (z) represents the value of the i-th membership function, n represents the total number of membership functions, and the weight coefficient is preset. According to the fuzzy membership value, a fuzzy control rule base based on temperature and current compensation is used to perform fuzzy reasoning operations, and the fuzzy control quantity of the charge and discharge current of each capacitor layer is generated by a fuzzy relationship mapper to achieve the mapping of the voltage difference and the internal resistance difference to the current control quantity. For the current fuzzy control quantity, a centroid method defuzzification unit is used to calculate the real-time charge and discharge current regulation quantity of each capacitor layer, and the amplitude of the regulation quantity is constrained by the current limit processor to generate a charge and discharge current instruction with amplitude limitation. Using the charge and discharge current instruction with amplitude limitation, a corresponding switch control signal is generated in the pulse width modulator, and real-time closed-loop regulation is performed according to the current detection signal of the plug-in charging module to achieve dynamic control of the charge and discharge current of each capacitor layer. The voltage and internal resistance detection device collected the voltage values ​​of the three-layer capacitors as 2.9 volts, 2.6 volts, and 2.5 volts, respectively. The voltage difference between adjacent layers was calculated to be 0.3 volts and 0.1 volts, and the internal resistance difference was 5 milliohms and 15 milliohms. The fuzzy variable processing unit divides the voltage difference into 7 language value intervals, negative large, negative medium, negative small, zero, positive small, positive medium, and positive large, and the corresponding domain range is -0.5 to 0.5 volts. The internal resistance difference is divided into 5 language value intervals, zero, small, medium, large, and extremely large, and the domain range is 0 to 20 milliohms. The fuzzy control rules for temperature and current compensation use a two-dimensional rule table, which contains 35 basic rules. At 25 degrees Celsius, when the voltage difference is 0.3 volts and the internal resistance difference is 5 milliohms, the corresponding rule output is a medium charging current. When the temperature rises to 75 degrees Celsius, due to the general decrease in internal resistance, the rule output is adjusted to low current charging under the same conditions. The fuzzy relationship mapping uses the Mamdan i reasoning method to map the membership values ​​of the voltage difference and the internal resistance difference to the charge and discharge current regulation amount.

[0035]

[0036] , μ out represents the membership value of the output, μV represents the membership function of the voltage difference, μR represents the membership function of the internal resistance difference, μI represents the membership function of the current regulation amount, and x i ,y i 、zi Respectively represent the corresponding input variable values. In the process of defuzzification by the centroid method, the fuzzy control quantity is converted into a specific current regulation value. The current limit processor sets the upper limit of the charging current to 2 amperes and the upper limit of the discharge current to 1.5 amperes. In actual operation, when the voltage difference is 0.3 volts, the calculated charging current regulation is 0.8 amperes, and when the voltage difference is reduced to 0.1 volts, the charging current is reduced to 0.3 amperes. For an internal resistance difference of 15 milliohms, the compensation current is set to 0.5 amperes. The switching frequency of the pulse width modulator is set to 20 kilohertz, and the dead time is 2 microseconds. The current detection uses a Hall sensor with a sampling frequency of 10 kilohertz and a detection accuracy of 0.01 amperes. In closed-loop control, when the actual charging current exceeds the command value by 10%, the controller adjusts the current to within the command range within 100 microseconds. It takes 5 seconds for the measured voltage difference to gradually decrease from 0.3 volts to 0.05 volts, and 10 seconds for the internal resistance difference to decrease from 15 milliohms to 8 milliohms. The entire process of fuzzy control reflects the comprehensive consideration of temperature and current. In low temperature environments, due to the generally large internal resistance, the charging current should not be set too large even if there is a large voltage difference. At high temperatures, the internal resistance decreases, resulting in improved charging efficiency. At this time, a smaller charging current can be used to achieve voltage balance. The system adjusts the charge and discharge current in real time, which not only ensures the voltage balancing speed, but also avoids the impact of excessive charge and discharge current on the life of the capacitor. In the dynamic adjustment process, the two key parameters of voltage and internal resistance restrict each other. A large internal resistance difference will lead to an increase in voltage imbalance during the charge and discharge process, and the voltage imbalance will affect the measurement accuracy of the internal resistance. Through the fuzzy control method, a nonlinear mapping relationship between voltage difference, internal resistance difference and charge and discharge current is established, and coordinated control of multiple mutually coupled parameters is achieved.

[0037] Step S105, during the emergency charging process based on the plug-in portable motorcycle, continuously monitor the temperature, current, voltage and internal resistance parameters of each capacitor layer of the plug-in charging module, and dynamically optimize the charge and discharge current of each capacitor layer according to the prediction results of the nonlinear mapping model and the target charge and discharge current control amount, so that the layered capacitor structure can achieve balanced charge and discharge.

[0038] The temperature, current, voltage and internal resistance parameters of the motorcycle are acquired during the charging process through the multi-channel collector of the plug-in charging module, and the temperature, current, voltage and internal resistance parameters are filtered and denoised by a parallel processor to obtain charging load characteristics; based on the charging load characteristics, a nonlinear mapping model is used to predict the capacitor layer parameters, and deviation data between the predicted value and the actual value of each capacitor layer parameter is generated; for the deviation data, the target current control amount is compared in the charge and discharge equalizer, and the compensation current value of each capacitor layer is calculated by the dynamic optimizer; based on the compensation current value, the charge and discharge current of each capacitor layer is closed-loop controlled in the equalization controller, and the controller is adjusted to a balanced charge and discharge state through a feedback regulator.

[0039] Specifically, the temperature, current, voltage, and internal resistance parameters of the portable motorcycle during charging are monitored in real time by the multi-channel collector of the plug-in charging module, and the collected data is filtered and denoised by a parallel processor, and the charging load characteristics and motor power characteristics are generated according to the motorcycle working condition identifier. In view of the charging load characteristics, a nonlinear mapping predictor is used to predict the parameters of the capacitor layer in real time, and the temperature, current, voltage, and internal resistance change rules are learned through a deep neural network to generate deviation data between the predicted value and the actual value of each capacitor layer parameter. According to the deviation data, the target current control amount is compared in the charge and discharge equalizer, and the compensation current value required for each capacitor layer is calculated by a dynamic optimizer, and the charge and discharge current of each capacitor layer is adjusted in real time by a current distributor. Based on the compensation current value, dynamic compensation is performed in the equalization controller according to the working state of the plug-in charging module, and the charge and discharge current of each capacitor layer is closed-loop controlled by a feedback regulator until a balanced charge and discharge state is reached. The multi-channel collector synchronously collects battery pack parameters during the charging process of the portable motorcycle, with a temperature sampling frequency of 10 Hz, a sampling accuracy of 0.1 degrees Celsius, and a measurement range of -20 to 85 degrees Celsius. The current sampling frequency is 1 kHz, the measurement accuracy is 0.01 ampere, and the range is 0 to 10 amperes. The motorcycle operating condition identifier divides the motor working state into four types: idle, light load, medium load, and heavy load, and the corresponding power loads are 100 watts, 300 watts, 500 watts, and 800 watts, respectively. The nonlinear mapping predictor is built based on a deep neural network. The input layer contains two parameters, temperature and current. The hidden layer adopts a 3-layer structure with 16 neurons in each layer. The output layer is the voltage and internal resistance prediction value. In actual application, when the motor is in a heavy load state and the power is 800 watts, the temperature of the capacitor layer increases from 25 degrees Celsius to 45 degrees Celsius, and the predicted internal resistance value decreases from 150 milliohms to 130 milliohms, and the deviation from the measured value is kept within 5%. The charge and discharge equalizer calculates the compensation current based on the deviation between the predicted value and the measured value. When the voltage deviation exceeds 0.1 volt, the compensation current is set to 0.5 amperes. The dynamic optimizer adopts an adaptive adjustment method. Under light load conditions, the compensation current is adjusted every 50 milliseconds, and the adjustment cycle is shortened to 20 milliseconds under heavy load conditions. The current distributor dynamically adjusts the charging current ratio of the three-layer capacitor to keep the maximum deviation within 10%. The balancing controller adopts a closed-loop control structure with a sampling period of 100 microseconds and a control period of 1 millisecond. When the plug-in charging module detects that the charging interface is connected, the controller completes initialization within 10 milliseconds. When the motor is idle, the controller limits the charging current to less than 1 ampere; when the motor enters a heavy load state, the upper limit of the charging current is increased to 2 amperes. The proportional coefficient and integral time of the feedback regulator are automatically adjusted according to the working conditions. A larger integral time is used under light load conditions to ensure control stability, and the integral time is reduced under heavy load conditions to improve the response speed. During the emergency charging process of portable motorcycles, changes in motor load directly affect the temperature distribution and internal resistance characteristics of the capacitor layer.When the road is flat and the motor is lightly loaded, the temperature of the capacitor layer is uniform, the internal resistance difference is small, and the charging and discharging control is mainly based on stability. When the motorcycle climbs a slope or accelerates quickly, the motor enters a heavy-load state, the temperature of the capacitor layer rises rapidly and is unevenly distributed, and the internal resistance shows a large difference. At this time, the controller prioritizes the balance of charging and discharging. During the control process, the four key parameters of temperature, current, voltage and internal resistance are coupled with each other. The increase in temperature causes the internal resistance to decrease and the charging and discharging efficiency to improve, but at the same time increases the parameter difference between the capacitor layers. High current charging and discharging aggravates the temperature non-uniformity and affects the voltage balance. Through real-time monitoring and prediction, combined with dynamic compensation for operating conditions, the capacitor layer can maintain a balanced charging and discharging state under various operating conditions. In practical applications, even when the motor power changes frequently, the voltage difference between the capacitor layers is controlled within 0.1 volts, and the internal resistance difference is controlled within 10 milliohms.

[0040] Step S106, during the emergency charging process based on the plug-in portable motorcycle, the motorcycle power is also monitored in real time, and an emergency charging plan for the portable motorcycle is generated based on the motorcycle power. If the motorcycle power is monitored to be in a preset low power range, the balancing control of the discharge of the layered capacitor structure of the plug-in charging module is triggered. If the monitored motorcycle power is higher than the preset high power threshold, the frequency of the balancing control is gradually reduced until the motorcycle power reaches full charge, and the balancing control is stopped and the charging is terminated.

[0041] The real-time power data of the motorcycle battery pack is obtained through a power monitor, and a power change curve is recorded according to the real-time power data; a comparison result between the power monitoring value and a preset threshold is generated according to the power change curve, and a power state judgment is made on the comparison result; if the power state judgment result is lower than the preset low power threshold, balanced discharge control is performed, and a neural network predictor is used to generate a discharge control parameter for a layered capacitor structure; according to the discharge control parameter, a power change trend is tracked in a charging state monitor, and if the power change trend reaches a preset full-charge state, a charging stop signal is generated and the charging circuit is disconnected.

[0042] Specifically, the real-time power of the motorcycle battery pack is collected through the power monitor, the power change curve is recorded by the data processing unit, the emergency charging strategy is formulated according to the charging scheme generator, and the comparison result of the power monitoring value and the threshold is generated. According to the comparison result, the power state is judged in the threshold judge. If the monitored power is lower than the preset low power threshold, the balanced discharge controller is started, and the discharge control parameters of the layered capacitor structure are generated by the neural network predictor. According to the discharge control parameters, the discharge balance frequency is set by the balanced controller, and the capacitor structure is discharged by the layered discharge unit. If the monitored power exceeds the preset high power threshold, the balanced control frequency is reduced according to the adaptive adjustment curve. Based on the adaptive adjustment curve, the power change trend is continuously tracked in the charging state monitor, and the motorcycle battery pack is determined to be fully charged according to the full charge judgement. The charging stop signal is generated by the charging terminator and the charging circuit is disconnected. The power monitor uses a high-precision power meter chip to monitor the motorcycle battery pack in real time, the sampling frequency is set to 10 Hz, and the power measurement accuracy is 0.1%. The power change curve recorded by the data processing unit is averaged with a time window of 10 seconds to eliminate the influence of instantaneous fluctuations. The charging scheme generator sets three charging strategies according to the power status: a fast charging scheme is used when the power is less than 20%, a standard charging scheme is used when the power is between 20% and 80%, and a trickle charging scheme is used when the power is above 80%. The threshold judge sets the low power threshold to 25% and the high power threshold to 85%. When the power is detected to be less than 25%, the neural network predictor generates discharge control parameters based on the motorcycle's driving conditions and ambient temperature. The predictor adopts a three-layer neural network structure. The input layer contains three parameters: power, temperature, and load power. The hidden layer contains 16 neurons. The output layer generates two control parameters: discharge current and discharge time. The balancing controller dynamically adjusts the discharge balancing frequency according to the power level. In the low power state, the balancing control frequency is set to 50 Hz to ensure fast response. The layered discharge unit adopts an alternating discharge method for the three-layer capacitor structure. The discharge time of each layer is 20 milliseconds and the switching time between layers is 2 milliseconds. When the power exceeds 85%, the adaptive adjustment curve reduces the control frequency according to the exponential function. For every 5% increase in power, the frequency is reduced to 0.7 times the original. The charging status monitor uses the coulomb meter method to accumulate the charge in real time with a monitoring accuracy of 1 mAh. The full charge judge defines the full charge state as the charging current drops to 0.05 times the rated value and lasts for more than 10 minutes. In actual application, when the motorcycle's power is 15%, emergency charging is started, and the layered capacitor is discharged at a current of 2 amperes, and the power rises at a rate of about 0.5% per minute. After the power reaches 30%, the discharge current drops to 1.5 amperes, and the rise rate slows down to 0.3% per minute. The power change process reflects the charging characteristics of different stages. In the low power stage, the charging speed is prioritized, and the capacitor layer adopts large current balanced discharge.As the power level increases, the charging speed gradually decreases, and more attention is paid to charging balance. When the power level is close to full, a small current is used to charge to avoid overcharging. Throughout the process, the dynamic adjustment of the balanced control frequency ensures a balance between charging efficiency and battery safety. Temperature has a significant impact on the charging process. At 25 degrees Celsius, the charging efficiency is the highest, and it only takes 40 minutes to charge from 20% to 80%. When the temperature drops to 0 degrees Celsius, the charging speed slows down by 50% due to the increase in the internal resistance of the battery. When the temperature rises to 45 degrees Celsius, although the charging speed increases, the controller will actively reduce the charging current, extend the charging time, and protect the battery. Through real-time monitoring and intelligent control, the system can ensure a safe and reliable charging process under various environmental conditions.

[0043] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the concept of the present application. For example, the above features are replaced with the technical features with similar functions disclosed in the present application (but not limited to) to form a technical solution.

Claims

1. A method for generating an emergency charging scheme for a portable motorcycle based on a plug-in design, characterized in that: The method comprises: Obtaining the capacitance voltage parameters and capacitance internal resistance parameters of each capacitance layer, taking the temperature and current data in the charging and discharging process as input, and taking the voltage and internal resistance parameters of each capacitance layer as output, training the nonlinear mapping model, and optimizing the capacitance, temperature and current parameters of the nonlinear mapping model through cross-validation; During the charging and discharging process of the plug-in charging module, the temperature and current data of each capacitor layer are collected in real time and input into the nonlinear mapping model to predict the voltage and internal resistance parameters of each capacitor layer to obtain the dynamic changes of the capacitor layer parameters during the charging and discharging process; The predicted voltage of each capacitor layer is compared with the preset voltage balance threshold to determine whether the voltage of each capacitor layer is balanced. If not, the voltage of each layer is adjusted through the energy transfer circuit until a balanced state is reached. The difference between the internal resistance of each capacitor layer and the predicted internal resistance parameter is calculated, and the internal resistance compensation value is determined according to the difference. The capacitor layer whose internal resistance difference is greater than the preset internal resistance difference is compensated; The calculated voltage difference and internal resistance difference of each capacitor layer are used as the input of the preset fuzzy control algorithm, and the charge and discharge current adjustment amount of each capacitor layer is used as the output. By fuzzifying the input variables, the target charge and discharge current control amount of each capacitor layer is output, and the target charge and discharge current control amount is transmitted to the current regulation circuit of the plug-in charging module to adjust the charge and discharge current of each capacitor layer in real time; During the emergency charging process of a plug-in portable motorcycle, the temperature, current, voltage and internal resistance parameters of each capacitor layer of the plug-in charging module are continuously monitored. According to the prediction results of the nonlinear mapping model and the target charge and discharge current control amount, the charge and discharge current of each capacitor layer is dynamically optimized to achieve balanced charge and discharge of the layered capacitor structure. During the emergency charging process based on the plug-in portable motorcycle, the motorcycle power is also monitored in real time, and an emergency charging plan for the portable motorcycle is generated based on the motorcycle power. If the motorcycle power is monitored to be in a preset low power range, the balancing control of the discharge of the layered capacitor structure of the plug-in charging module is triggered. If the monitored motorcycle power is higher than the preset high power threshold, the frequency of balancing control is gradually reduced until the motorcycle power reaches full charge, and the balancing control is stopped and charging is terminated.

2. The method according to claim 1, characterized in that: The method of obtaining the capacitance voltage parameters and capacitance internal resistance parameters of each capacitance layer, taking the temperature and current data in the charging and discharging process as input, taking the voltage and internal resistance parameters of each capacitance layer as output, training the nonlinear mapping model, and optimizing the capacitance, temperature and current parameters of the nonlinear mapping model through cross-validation includes: Acquire temperature signals and current signals during the charging and discharging process from a parameter acquisition device, wherein the parameter acquisition device records voltage values ​​and internal resistance values ​​at corresponding moments; According to the temperature signal and the current signal, removing abnormal fluctuation points through a data cleaning unit to obtain a standardized training data set; For the standardized training data set, cross-grouping is performed using a data partitioning unit, and a training data subset and a validation data subset are generated according to the proportion of the training set; Establishing a nonlinear mapping function from the temperature signal and the current signal to the voltage value and the internal resistance value through a neural network mapper, and training the parameters of the nonlinear mapping function using a training data subset to obtain a first round of optimization parameter set; If the average error value on the validation data subset is higher than the average error threshold, the first round optimization parameter set is iteratively updated using a gradient descent optimizer until the average error value is lower than the average error threshold.

3. The method according to claim 1, characterized in that During the charging and discharging process of the plug-in charging module, the temperature and current data of each capacitor layer are collected in real time, input into the nonlinear mapping model, and the voltage and internal resistance parameters of each capacitor layer are predicted to obtain the dynamic changes of the capacitor layer parameters during the charging and discharging process, including: Receiving the surface temperature distribution data and current change data of the capacitor layer collected by the plug-in charging module, wherein the temperature distribution data and the current change data are generated into a real-time monitoring data set with a time stamp according to a fixed sampling time interval; A data preprocessing unit is used to perform standardization processing on the real-time monitoring data set, and a nonlinear mapping model is used to predict the voltage parameter and internal resistance parameter of the capacitor layer according to the standardized data; Extracting time series features according to the predicted voltage parameters and internal resistance parameters, and the state identifier generating a time series correspondence between the voltage parameters, the internal resistance parameters, the temperature data, and the current data according to the time series features; A Kalman filter is used to smooth the voltage parameters and internal resistance parameters in the time sequence correspondence, and the smoothed parameters are combined with the working state of the plug-in charging module to obtain a dynamic change curve of the capacitance layer parameters.

4. The method according to claim 1, characterized in that The predicted voltage of each capacitor layer is compared with a preset voltage balance threshold to determine whether the voltage of each capacitor layer is balanced. If not, the voltage of each layer is adjusted by an energy transfer circuit until a balanced state is reached, the difference between the internal resistance of each capacitor layer and the predicted internal resistance parameter is calculated, and the internal resistance compensation value is determined according to the difference. The capacitor layer whose internal resistance difference is greater than the preset internal resistance difference is compensated, including: A voltage detector is used to compare and calculate the predicted voltage value of the capacitor layer with the preset voltage balance threshold to obtain the voltage deviation and the inter-layer voltage difference value; According to the comparison result between the voltage deviation and the preset threshold range, an energy transfer channel is established between adjacent capacitor layers through an energy transfer controller; For the energy transfer channel, the energy transfer time and the energy transfer current are set according to the interlayer voltage difference value, and the internal resistance difference value of each capacitor layer is calculated by an internal resistance comparator; For the internal resistance difference value, if it exceeds the preset difference value, the compensation parameter generator generates the internal resistance compensation parameter, and the internal resistance compensation process is tracked in real time through the Kalman filter until the internal resistance difference value drops below the preset difference value.

5. The method according to claim 1, characterized in that The calculated voltage difference and internal resistance difference of each capacitor layer are used as inputs of a preset fuzzy control algorithm, and the charge and discharge current adjustment amount of each capacitor layer is used as output. By fuzzifying the input variables, the target charge and discharge current control amount of each capacitor layer is output, and the target charge and discharge current control amount is transmitted to the current adjustment circuit of the plug-in charging module, and the charge and discharge current of each capacitor layer is adjusted in real time, including: The voltage difference and internal resistance difference between each capacitor layer are obtained by a voltage internal resistance detection device, and a fuzzy variable processing unit is used to generate a fuzzy membership value for the voltage difference and the internal resistance difference; According to the fuzzy membership value, a fuzzy control rule base of temperature and current compensation is used to perform fuzzy reasoning operation to obtain the fuzzy control amount of the charge and discharge current of each capacitor layer; For the current fuzzy control amount, a centroid method defuzzification unit is used to calculate the real-time charge and discharge current regulation amount of each capacitor layer, and a charge and discharge current instruction with amplitude limitation is generated according to the regulation amount; The charge and discharge current instruction with amplitude limitation is adopted to generate a switch control signal in a pulse width modulator, and the charge and discharge current is adjusted according to the current detection signal of the plug-in charging module.

6. The method according to claim 1, characterized in that In the process of emergency charging based on the plug-in portable motorcycle, the temperature, current, voltage and internal resistance parameters of each capacitor layer of the plug-in charging module are continuously monitored, and the charge and discharge current of each capacitor layer is dynamically optimized according to the prediction result of the nonlinear mapping model and the target charge and discharge current control amount, so that the layered capacitor structure can achieve balanced charge and discharge, including: The temperature, current, voltage and internal resistance parameters of the motorcycle during charging are obtained through a multi-channel collector of the plug-in charging module, and the temperature, current, voltage and internal resistance parameters are filtered and de-noised by a parallel processor to obtain charging load characteristics; According to the charging load characteristics, a nonlinear mapping model is used to predict the capacitance layer parameters, and deviation data between the predicted value and the actual value of each capacitance layer parameter is generated; For the deviation data, the target current control amount is compared in the charge and discharge equalizer, and the compensation current value of each capacitor layer is calculated by the dynamic optimizer; According to the compensation current value, closed-loop control is performed on the charge and discharge current of each capacitor layer in the balance controller, and the controller is adjusted to a balanced charge and discharge state through a feedback regulator.

7. The method according to claim 1, characterized in that In the process of emergency charging based on the plug-in portable motorcycle, the motorcycle power is also monitored in real time, and a portable motorcycle emergency charging plan is generated according to the motorcycle power. If the motorcycle power is monitored to be within a preset low power range, the balanced control of the discharge of the layered capacitor structure of the plug-in charging module is triggered. If the monitored motorcycle power is higher than the preset high power threshold, the frequency of the balanced control is gradually reduced until the motorcycle power reaches full power, the balanced control is stopped and the charging is terminated, including: Acquire real-time power data of the motorcycle battery pack through a power monitor, and record a power change curve according to the real-time power data; Generate a comparison result between the power monitoring value and a preset threshold value according to the power change curve, and perform power state determination on the comparison result; If the state of charge determination result is lower than a preset low-power threshold, balanced discharge control is performed, and a neural network predictor is used to generate discharge control parameters for a layered capacitor structure; The charge change trend is tracked in the charging state monitor according to the discharge control parameter. If the charge change trend reaches a preset full charge state, a charging stop signal is generated and the charging circuit is disconnected.

Images