A method for intelligent recovery control of braking energy in electric vehicles

Abstract

This invention discloses an intelligent control method for regenerative braking energy in electric vehicles, comprising: condition-specific control based on braking intensity; calculating the target torque according to the driving mode under low braking intensity; calculating the vehicle stability safety coefficient by combining braking pressure, wheel speed, and deceleration under medium-to-high braking intensity; dynamically allocating motor torque and hydraulic braking torque; and further calculating the maximum allowable torque. During testing, data on braking pressure, wheel speed, deceleration, and charging power are collected; a particle swarm optimization algorithm is introduced to predict the energy recovery trend; and the relationship between the trend and vehicle state is evaluated to optimize the kinetic energy recovery strategy. This invention enables the calculation of the maximum allowable torque for electric vehicles during kinetic energy recovery, improving the energy recovery efficiency of new energy vehicles while enhancing battery safety. It also achieves accurate prediction and evaluation of kinetic energy recovery effects, providing data support for developing efficient recovery strategies.

Description

Technical Field

[0001] This invention relates to the field of kinetic energy analysis for electric vehicles, and more specifically, to a method for intelligent recovery control of braking energy in electric vehicles. Background Technology

[0002] With the booming development of new energy vehicles, regenerative braking systems, as a key technology for improving vehicle energy utilization efficiency, have been widely applied. Most existing regenerative braking systems for new energy vehicles limit the regenerative power based on the battery's state of charge (SOC) threshold, outputting braking torque based on the efficiency of the drive motor. However, this traditional approach has revealed several problems in practical applications. First, when the efficiency of the drive motor system in a new energy vehicle deviates from the theoretical value, the regenerative torque calculated based on the SOC threshold can easily cause instantaneous overcharging of the vehicle battery, leading to the risk of battery voltage exceeding limits, battery damage, and reduced battery life. Second, to prevent battery overcharging, existing technologies often limit the regenerative braking power. However, this approach results in the loss of a large amount of recoverable energy, reducing the energy recovery efficiency of new energy vehicles and contradicting the original intention of energy conservation and environmental protection in new energy vehicles. Simultaneously, existing technologies lack methods for assessing the correlation between the energy recovery power trend of electric vehicles and the vehicle's state, resulting in inefficient recovery control and insufficient adaptability. Summary of the Invention

[0003] This invention overcomes the shortcomings of the prior art and proposes a method for intelligent recovery control of braking energy in electric vehicles.

[0004] The first aspect of this invention provides a method for intelligent regenerative braking control of electric vehicles, comprising: S1: During a test period, the target electric vehicle is powered on for testing to determine whether the vehicle is in the expected state of kinetic energy recovery. If so, the braking intensity is calculated in real time. S2: When the braking intensity is low, the VCU selects the target energy recovery level according to the driving mode, calculates the target deceleration according to the vehicle speed, performs PI control by comparing the target deceleration with the actual vehicle acceleration value issued by the current ABS, and calculates the target torque of the motor. S3: When the braking intensity is medium to high, the VCU calculates the vehicle's stability safety factor based on the braking pressure, wheel speed, and deceleration parameters, and calculates the total braking torque. It then subtracts the real-time motor torque from the total braking torque and uses the calculation results to feed back and control the hydraulic braking torque. S4: After comprehensively comparing the maximum energy recovery torque, braking torque, and maximum allowable torque, the VCU outputs the torque control result to the MCU for motor control; S5: Divide a test period into multiple time points; S6: Analyze the numerical changes of braking pressure, wheel speed, and deceleration parameters during the test period, generate parameter change curves, generate energy recovery change curves based on the charging power of the kinetic energy recovery process, introduce particle swarm optimization algorithm, use braking pressure, wheel speed, and deceleration as decision variables, set parameter constraints, use maximizing energy recovery as the objective function to search for local optimal parameters, and select the preferred time point based on the time point where the local optimal parameters are located. S7: Select the corresponding energy recovery data by optimizing the time points, serialize the energy recovery data, perform sequence prediction through the prediction model, evaluate the variable relationship between energy recovery trend and electric vehicle state, and optimize the kinetic energy recovery strategy.

[0005] In this solution, the expected state includes: A high-voltage power-on test was conducted on the target electric vehicle. The vehicle was in the Ready state, in the D gear, and the accelerator pedal was released.

[0006] In this scheme, the maximum permissible torque is calculated as follows: Collect information on the voltage and temperature of individual battery cells in the target electric vehicle, and calculate the maximum allowable charging power of the battery; The actual battery power is calculated by collecting the actual battery pack voltage and total output current of the target electric vehicle. Calculate the charging power deviation as follows: Actual battery power - Maximum allowable battery charging power; When the charging power deviation is less than the set value of 1: Set the power rate decrease value; calculate the maximum allowable charging power = maximum allowable battery charging power - (power rate decrease * dt); When the charging power deviation is less than the set value of 2: Set the incremental power rate value; calculate the maximum allowable charging power = maximum allowable battery charging power + (incremental power rate * dt); otherwise: Maximum allowable charging power = maximum allowable charging power of the battery; Calculate the maximum regenerative braking power = maximum allowable charging power + accessory power; Calculate the maximum allowable torque for regenerative braking = maximum regenerative braking power * 9550 / current motor speed.

[0007] In this scheme, in step S5, for a test period, a preset time interval is set to divide multiple time points, and the interval between adjacent time points remains consistent.

[0008] In this solution, S6 specifically refers to: During the test period, numerical changes in braking pressure, wheel speed, and deceleration parameters were analyzed, and curves showing the changes in these three parameters were generated. Based on the change analysis of charging power during the kinetic energy recovery process, an energy recovery change curve is generated; A particle swarm optimization algorithm is introduced, the number of particles M is set, and each particle corresponds to a combination of three parameters. Using braking pressure, wheel speed, and deceleration as decision variables, parameter constraints are set based on the maximum theoretical parameters of the vehicle state. In the parameter constraints, add coupling constraints, specifically using the fact that braking pressure and wheel speed are positively correlated with deceleration as constraints. In the energy recovery change curve, the goal function is to maximize the charging power. A local optimum search is performed on all particles, and the local optimum parameter combination is obtained based on the search results. In the local optimal parameter combination, the local optimal time T is marked by the parameter change curve, and the nearest time point is selected from multiple time points based on time T and marked as the preferred time point.

[0009] In this scheme, S6 also includes multiple local optimal times T.

[0010] In this solution, S7 specifically refers to: Construct a prediction model based on ARMA; Based on the preferred time points, energy recovery data is extracted from the energy recovery change curve; The energy recovery data is time-series converted to form an energy recovery sequence; The energy recovery sequence was verified for stationarity using ADF, and the non-stationary parts were processed by differential processing. Calculate the autocorrelation function and partial autocorrelation function in the energy recovery change curve, and determine the parameters p and q; The prediction model is fitted using the maximum likelihood estimation method, and the energy recovery sequence is predicted based on the prediction model to obtain the predicted sequence. By predicting the sequence, the variable relationship between energy recovery trends and electric vehicle status is evaluated, and the kinetic energy recovery strategy is optimized.

[0011] In this solution, S7 further includes: Linear regression was performed to fit the parameter variation curves, and the goodness of fit of the parameter variation curves was analyzed. ; judge Is it greater than the preset goodness? If so, the energy recovery sequence is predicted by linear regression, with a prediction step size of N time points, to obtain the predicted sequence; Combination By comparing the predicted sequence with the variables of energy recovery trend and electric vehicle status, we can assess the relationship between these variables.

[0012] A second aspect of the present invention also provides an intelligent regenerative braking control system for electric vehicles. The system includes: a memory, a processor, and a data interface / communication interface. The memory includes an intelligent regenerative braking control program for electric vehicles. When executed by the processor, the intelligent regenerative braking control program for electric vehicles performs the following steps: S1: During a test period, the target electric vehicle is powered on for testing to determine whether the vehicle is in the expected state of kinetic energy recovery. If so, the braking intensity is calculated in real time. S2: When the braking intensity is low, the VCU selects the target energy recovery level according to the driving mode, calculates the target deceleration according to the vehicle speed, performs PI control by comparing the target deceleration with the actual vehicle acceleration value issued by the current ABS, and calculates the target torque of the motor. S3: When the braking intensity is medium to high, the VCU calculates the vehicle's stability safety factor based on the braking pressure, wheel speed, and deceleration parameters, and calculates the total braking torque. It then subtracts the real-time motor torque from the total braking torque and uses the calculation results to feed back and control the hydraulic braking torque. S4: After comprehensively comparing the maximum energy recovery torque, braking torque, and maximum allowable torque, the VCU outputs the torque control result to the MCU for motor control; S5: Divide a test period into multiple time points; S6: Analyze the numerical changes of braking pressure, wheel speed, and deceleration parameters during the test period, generate parameter change curves, generate energy recovery change curves based on the charging power of the kinetic energy recovery process, introduce particle swarm optimization algorithm, use braking pressure, wheel speed, and deceleration as decision variables, set parameter constraints, use maximizing energy recovery as the objective function to search for local optimal parameters, and select the preferred time point based on the time point where the local optimal parameters are located. S7: Select the corresponding energy recovery data by optimizing the time points, serialize the energy recovery data, perform sequence prediction through the prediction model, evaluate the variable relationship between energy recovery trend and electric vehicle state, and optimize the kinetic energy recovery strategy.

[0013] A third aspect of the present invention also provides a computer-readable storage medium comprising an intelligent recovery control program for electric vehicle braking energy, wherein when executed by a processor, the intelligent recovery control program for electric vehicle braking energy implements the steps of the intelligent recovery control method for electric vehicle braking energy as described in any of the preceding claims.

[0014] This invention discloses an intelligent control method for regenerative braking energy in electric vehicles, comprising: condition-specific control based on braking intensity; calculating the target torque according to the driving mode under low braking intensity; calculating the vehicle stability safety coefficient by combining braking pressure, wheel speed, and deceleration under medium-to-high braking intensity; dynamically allocating motor torque and hydraulic braking torque; and further calculating the maximum allowable torque. During testing, data on braking pressure, wheel speed, deceleration, and charging power are collected; a particle swarm optimization algorithm is introduced to predict the energy recovery trend; and the relationship between the trend and vehicle state is evaluated to optimize the kinetic energy recovery strategy. This invention enables the calculation of the maximum allowable torque for electric vehicles during kinetic energy recovery, improving the energy recovery efficiency of new energy vehicles while enhancing battery safety. It also achieves accurate prediction and evaluation of kinetic energy recovery effects, providing data support for developing efficient recovery strategies. Attached Figure Description

[0015] Figure 1 A flowchart illustrating the calculation of braking intensity and torque according to the present invention is shown; Figure 2 A flowchart illustrating the calculation of the maximum permissible torque according to the present invention is shown; Figure 3 A flowchart of the energy recovery prediction process of the present invention is shown; Figure 4 A block diagram of an intelligent regenerative braking control system for electric vehicles according to the present invention is shown. Detailed Implementation

[0016] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It is understood that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0017] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0018] Figure 1 A flowchart illustrating the calculation of braking intensity and torque according to the present invention is shown.

[0019] The first aspect of this invention provides a method for intelligent regenerative braking control of electric vehicles, comprising: S1: During a test period, the target electric vehicle is powered on for testing to determine whether the vehicle is in the expected state of kinetic energy recovery. If so, the braking intensity is calculated in real time. S2: When the braking intensity is low, the VCU selects the target energy recovery level according to the driving mode, calculates the target deceleration according to the vehicle speed, performs PI control by comparing the target deceleration with the actual vehicle acceleration value issued by the current ABS, and calculates the target torque of the motor. S3: When the braking intensity is medium to high, the VCU calculates the vehicle's stability safety factor based on the braking pressure, wheel speed, and deceleration parameters, and calculates the total braking torque. It then subtracts the real-time motor torque from the total braking torque and uses the calculation results to feed back and control the hydraulic braking torque. S4: After comprehensively comparing the maximum energy recovery torque, braking torque, and maximum allowable torque, the VCU outputs the torque control result to the MCU for motor control; S5: Divide a test period into multiple time points; S6: Analyze the numerical changes of braking pressure, wheel speed, and deceleration parameters during the test period, generate parameter change curves, generate energy recovery change curves based on the charging power of the kinetic energy recovery process, introduce particle swarm optimization algorithm, use braking pressure, wheel speed, and deceleration as decision variables, set parameter constraints, use maximizing energy recovery as the objective function to search for local optimal parameters, and select the preferred time point based on the time point where the local optimal parameters are located. S7: Select the corresponding energy recovery data by optimizing the time points, serialize the energy recovery data, perform sequence prediction through the prediction model, evaluate the variable relationship between energy recovery trend and electric vehicle state, and optimize the kinetic energy recovery strategy.

[0020] It can be understood here that new energy vehicles include various controllers, such as: vehicle control unit (VCU), engine control unit (ECU), motor control unit (MCU), and battery management system (BMS).

[0021] With the control method of this invention, when the actual efficiency of the motor is low, after the VCU target braking recovery output torque is executed, the charging power will be greater than the theoretical power. At this time, the actual power of the battery will be greater than the maximum allowable charging power of the battery, resulting in a power deviation. Through this power deviation, the maximum allowable charging power is reduced, thereby reducing the maximum allowable charging torque. The target torque output by the VCU to the motor will be lower than that of the previous operating cycle, thereby reducing the charging power, preventing the battery from overcharging, and effectively utilizing the recoverable kinetic energy.

[0022] According to an embodiment of the present invention, the expected state includes: A high-voltage power-on test was conducted on the target electric vehicle. The vehicle was in the Ready state, in the D gear, and the accelerator pedal was released.

[0023] According to an embodiment of the present invention, the maximum permissible torque is calculated as follows: Collect information on the voltage and temperature of individual battery cells in the target electric vehicle, and calculate the maximum allowable charging power of the battery; The actual battery power is calculated by collecting the actual battery pack voltage and total output current of the target electric vehicle. Calculate the charging power deviation as follows: Actual battery power - Maximum allowable battery charging power; When the charging power deviation is less than the set value of 1: Set the power rate decrease value; calculate the maximum allowable charging power = maximum allowable battery charging power - (power rate decrease * dt); When the charging power deviation is less than the set value of 2: Set the incremental power rate value; calculate the maximum allowable charging power = maximum allowable battery charging power + (incremental power rate * dt); otherwise: Maximum allowable charging power = maximum allowable charging power of the battery; Calculate the maximum regenerative braking power = maximum allowable charging power + accessory power; Calculate the maximum allowable torque for regenerative braking = maximum regenerative braking power * 9550 / current motor speed.

[0024] It is understandable that the decreasing / increasing power rate * dt represents the cumulative power over time. The decreasing / increasing power rate value is set by the user, usually based on the rated power of the electric vehicle.

[0025] According to an embodiment of the present invention, in step S5, for a test period, a preset time interval is set to divide multiple time points, and the interval between adjacent time points remains consistent.

[0026] In actual vehicle power-on testing, due to the influence of testing equipment, testing schemes, sensor errors, etc., the various parameters collected will have certain error values ​​in different driving modes. Specifically, the data will have certain noise. In particular, for complex driving environments and different driving habits, the corresponding vehicle acceleration and braking situations are more complex. In the process of kinetic energy recovery braking control and kinetic energy recovery prediction analysis, the calculation and analysis process needs to optimize and extract data in order to effectively achieve accurate analysis and evaluation of kinetic energy recovery.

[0027] This invention effectively solves the aforementioned problems and provides a method for efficient kinetic energy trend analysis and joint evaluation of vehicle status. This invention analyzes the changing characteristics of braking pressure, wheel speed, and deceleration parameters, representing them as curves or functions. It introduces a particle swarm optimization algorithm to perform a local optimum search during the test period, aiming to maximize energy recovery. This local optimum search filters out time periods that avoid noise. Energy recovery data is extracted using these optimized time points, and based on the changing characteristics of kinetic energy recovery-related curves, the trend of power in energy recovery is predicted using an ARMA model or linear regression. By combining the predicted data with the vehicle's status parameters, the kinetic energy recovery status and efficiency of the vehicle can be determined.

[0028] By combining the predicted sequence with the vehicle state parameters in this invention, the effectiveness of the energy recovery strategy under different driving modes, the trend of energy recovery, and the potential relationship between different state modes and energy trends can be accurately evaluated. The monitoring modes and energy recovery strategies of the VCU and MCU modules can be adjusted, thereby enabling efficient testing and accurate simulation analysis of energy recovery for electric vehicles.

[0029] According to an embodiment of the present invention, S6 specifically includes: During the test period, numerical changes in braking pressure, wheel speed, and deceleration parameters were analyzed, and curves showing the changes in these three parameters were generated. Based on the change analysis of charging power during the kinetic energy recovery process, an energy recovery change curve is generated; A particle swarm optimization algorithm is introduced, the number of particles M is set, and each particle corresponds to a combination of three parameters. Using braking pressure, wheel speed, and deceleration as decision variables, parameter constraints are set based on the maximum theoretical parameters of the vehicle state. In the parameter constraints, add coupling constraints, specifically using the fact that braking pressure and wheel speed are positively correlated with deceleration as constraints. In the energy recovery change curve, the goal function is to maximize the charging power. A local optimum search is performed on all particles, and the local optimum parameter combination is obtained based on the search results. In the local optimal parameter combination, the local optimal time T is marked by the parameter change curve, and the nearest time point is selected from multiple time points based on time T and marked as the preferred time point.

[0030] Understandably, in parameter variation curves, the variation curve (function) can be fitted using linear regression, quadratic fitting, exponential fitting, etc., and braking pressure, wheel speed, and deceleration often exhibit certain linear regression variation patterns. Each "particle" is composed of three dimensions of data [braking pressure, wheel speed, deceleration]. Local optimum search can generally find multiple combinations of results. With maximizing charging power as the objective function, the local optimum search range can be set to the top 10% of maximum charging power. Multiple values ​​of T exist. Energy recovery power is generally 0-100kW.

[0031] The decision variables include braking pressure (MPa), four wheel speeds (km / h, which can be averaged or minimized), and vehicle deceleration (m / s²). The parameter variation curves can be visualized using Python's Matplotlib. Specifically, the three parameter curves are the braking pressure-time curve, the wheel speed-time curve, and the deceleration-time curve. Charging power is the actual battery power, used here to reflect the actual kinetic energy recovery. The locally optimal parameter combination is represented as [P_opt, V_opt, A_opt], ​​where each represents a locally optimal solution for braking pressure, wheel speed, and deceleration, respectively.

[0032] According to an embodiment of the present invention, S6 further includes a plurality of local optimal times T.

[0033] According to an embodiment of the present invention, S7 specifically includes: Construct a prediction model based on ARMA; Based on the preferred time points, energy recovery data is extracted from the energy recovery change curve; The energy recovery data is time-series converted to form an energy recovery sequence; The energy recovery sequence was verified for stationarity using ADF, and the non-stationary parts were processed by differential processing. Calculate the autocorrelation function and partial autocorrelation function in the energy recovery change curve, and determine the parameters p and q; The prediction model is fitted using the maximum likelihood estimation method, and the energy recovery sequence is predicted based on the prediction model to obtain the predicted sequence. By predicting the sequence, the variable relationship between energy recovery trends and electric vehicle status is evaluated, and the kinetic energy recovery strategy is optimized.

[0034] It is understood that the energy recovery data includes multiple charging power values. This invention uses the ARMA model for stationary sequence prediction, which is reflected in the predictive analysis and trend change analysis of energy recovery efficiency. The autocorrelation function (ACF) and partial autocorrelation function (PACF) are the core steps in constructing the ARMA model. Here, due to the uncertainty in the number of selected preferred time points, and the high correlation between the change characteristics of the energy recovery sequence and the energy recovery change curve, the energy recovery change curve is used as the raw data for autocorrelation function and partial autocorrelation function analysis. This improves the analysis process of the traditional ARMA model and determines the p and q parameters based on this, rather than using existing sequences for correlation function analysis. This can predict the non-stationary change characteristics of energy recovery, which is helpful for subsequent correlation assessment of energy recovery trends and vehicle state factors.

[0035] Parameter determination includes: determining the order (p) of the autoregressive component and the order (q) of the moving average component.

[0036] The electric vehicle status can include information such as braking pressure, wheel speed, deceleration, braking intensity, and braking torque. Based on the changes in vehicle status and the predicted trend of energy recovery status, the energy recovery trend of different driving modes is correlated and evaluated, and the optimal driving mode in local time is dynamically selected. In addition, by combining the analysis of vehicle braking intensity and the calculation of the maximum allowable charging power of the energy recovery process by MCU and VCU, the control period and control parameters can be optimized, thereby realizing an adaptive intelligent system that dynamically finds and maintains the optimal energy recovery point, providing effective data support for the research on kinetic energy recovery of electric vehicles.

[0037] According to an embodiment of the present invention, S7 further includes: Linear regression was performed to fit the parameter variation curves, and the goodness of fit of the parameter variation curves was analyzed. ; judge Is it greater than the preset goodness? If so, the energy recovery sequence is predicted by linear regression, with a prediction step size of N time points, to obtain the predicted sequence; Combination By comparing the predicted sequence with the variables of energy recovery trend and electric vehicle status, we can assess the relationship between these variables.

[0038] Understandably, for kinetic energy recovery test cycles with a certain degree of linear variation, linear regression can be used for fitting and sequence prediction, and the sequence prediction effect is better than that of the ARMA model.

[0039] The parameter variation curve can be evaluated using one or more parameters. Generally speaking, The goodness of fit is based on the average goodness of fit of the curves showing the variation of three parameters. Goodness of fit reflects both the linearity of kinetic energy recovery and its predictability and effectiveness. The larger the value, the greater the predictability and predictive effectiveness. The prediction step size is specifically the length formed by N consecutive time points.

[0040] Figure 2 A flowchart illustrating the calculation of the maximum permissible torque according to the present invention is shown; Figure 3 A flowchart of the energy recovery prediction process of the present invention is shown; Figure 4 A block diagram of an intelligent regenerative braking control system for electric vehicles according to the present invention is shown.

[0041] A second aspect of the present invention also provides an intelligent regenerative braking control system for electric vehicles. The system includes: a memory U3, a processor U2, and a data interface U1. The data interface is used to feed back parameter data calculated during a test period to the vehicle terminal module. The memory includes an intelligent regenerative braking control program for electric vehicles. When the processor executes the intelligent regenerative braking control program, it performs the following steps: S1: During a test period, the target electric vehicle is powered on for testing to determine whether the vehicle is in the expected state of kinetic energy recovery. If so, the braking intensity is calculated in real time. S2: When the braking intensity is low, the VCU selects the target energy recovery level according to the driving mode, calculates the target deceleration according to the vehicle speed, performs PI control by comparing the target deceleration with the actual vehicle acceleration value issued by the current ABS, and calculates the target torque of the motor. S3: When the braking intensity is medium to high, the VCU calculates the vehicle's stability safety factor based on the braking pressure, wheel speed, and deceleration parameters, and calculates the total braking torque. It then subtracts the real-time motor torque from the total braking torque and uses the calculation results to feed back and control the hydraulic braking torque. S4: After comprehensively comparing the maximum energy recovery torque, braking torque, and maximum allowable torque, the VCU outputs the torque control result to the MCU for motor control; S5: Divide a test period into multiple time points; S6: Analyze the numerical changes of braking pressure, wheel speed, and deceleration parameters during the test period, generate parameter change curves, generate energy recovery change curves based on the charging power of the kinetic energy recovery process, introduce particle swarm optimization algorithm, use braking pressure, wheel speed, and deceleration as decision variables, set parameter constraints, use maximizing energy recovery as the objective function to search for local optimal parameters, and select the preferred time point based on the time point where the local optimal parameters are located. S7: Select the corresponding energy recovery data by optimizing the time points, serialize the energy recovery data, perform sequence prediction through the prediction model, evaluate the variable relationship between energy recovery trend and electric vehicle state, and optimize the kinetic energy recovery strategy.

[0042] It can be understood here that new energy vehicles include various controllers, such as: vehicle control unit (VCU), engine control unit (ECU), motor control unit (MCU), and battery management system (BMS).

[0043] According to an embodiment of the present invention, the expected state includes: A high-voltage power-on test was conducted on the target electric vehicle. The vehicle was in the Ready state, in the D gear, and the accelerator pedal was released.

[0044] According to an embodiment of the present invention, the maximum permissible torque is calculated as follows: Collect information on the voltage and temperature of individual battery cells in the target electric vehicle, and calculate the maximum allowable charging power of the battery; The actual battery power is calculated by collecting the actual battery pack voltage and total output current of the target electric vehicle. Calculate the charging power deviation as follows: Actual battery power - Maximum allowable battery charging power; When the charging power deviation is less than the set value of 1: Set the power rate decrease value; calculate the maximum allowable charging power = maximum allowable battery charging power - (power rate decrease * dt); When the charging power deviation is less than the set value of 2: Set the incremental power rate value; calculate the maximum allowable charging power = maximum allowable battery charging power + (incremental power rate * dt); otherwise: Maximum allowable charging power = maximum allowable charging power of the battery; Calculate the maximum regenerative braking power = maximum allowable charging power + accessory power; Calculate the maximum allowable torque for regenerative braking = maximum regenerative braking power * 9550 / current motor speed.

[0045] It is understandable that the decreasing / increasing power rate * dt represents the cumulative power over time. The decreasing / increasing power rate value is set by the user, usually based on the rated power of the electric vehicle.

[0046] According to an embodiment of the present invention, in step S5, for a test period, a preset time interval is set to divide multiple time points, and the interval between adjacent time points remains consistent.

[0047] In actual vehicle power-on testing, due to the influence of testing equipment, testing schemes, sensor errors, etc., the various parameters collected will have certain error values ​​in different driving modes. Specifically, the data will have certain noise. In particular, for complex driving environments and different driving habits, the corresponding vehicle acceleration and braking situations are more complex. In the process of kinetic energy recovery braking control and kinetic energy recovery prediction analysis, the calculation and analysis process needs to optimize and extract data in order to effectively achieve accurate analysis and evaluation of kinetic energy recovery.

[0048] According to an embodiment of the present invention, S6 specifically includes: During the test period, numerical changes in braking pressure, wheel speed, and deceleration parameters were analyzed, and curves showing the changes in these three parameters were generated. Based on the change analysis of charging power during the kinetic energy recovery process, an energy recovery change curve is generated; A particle swarm optimization algorithm is introduced, the number of particles M is set, and each particle corresponds to a combination of three parameters. Using braking pressure, wheel speed, and deceleration as decision variables, parameter constraints are set based on the maximum theoretical parameters of the vehicle state. In the parameter constraints, add coupling constraints, specifically using the fact that braking pressure and wheel speed are positively correlated with deceleration as constraints. In the energy recovery change curve, the goal function is to maximize the charging power. A local optimum search is performed on all particles, and the local optimum parameter combination is obtained based on the search results. In the local optimal parameter combination, the local optimal time T is marked by the parameter change curve, and the nearest time point is selected from multiple time points based on time T and marked as the preferred time point.

[0049] Understandably, in parameter variation curves, the variation curve (function) can be fitted using linear regression, quadratic fitting, exponential fitting, etc., and braking pressure, wheel speed, and deceleration often exhibit certain linear regression variation patterns. Each "particle" is composed of three dimensions of data [braking pressure, wheel speed, deceleration]. Local optimum search can generally find multiple combinations of results. With maximizing charging power as the objective function, the local optimum search range can be set to the top 10% of maximum charging power. Multiple values ​​of T exist. Energy recovery power is generally 0-100kW.

[0050] The decision variables include braking pressure (MPa), four wheel speeds (km / h, which can be averaged or minimized), and vehicle deceleration (m / s²). The parameter variation curves can be visualized using Python's Matplotlib. Specifically, the three parameter curves are the braking pressure-time curve, the wheel speed-time curve, and the deceleration-time curve. Charging power is the actual battery power, used here to reflect the actual kinetic energy recovery. The locally optimal parameter combination is represented as [P_opt, V_opt, A_opt], ​​where each represents a locally optimal solution for braking pressure, wheel speed, and deceleration, respectively.

[0051] According to an embodiment of the present invention, S6 further includes a plurality of local optimal times T.

[0052] According to an embodiment of the present invention, S7 specifically includes: Construct a prediction model based on ARMA; Based on the preferred time points, energy recovery data is extracted from the energy recovery change curve; The energy recovery data is time-series converted to form an energy recovery sequence; The energy recovery sequence was verified for stationarity using ADF, and the non-stationary parts were processed by differential processing. Calculate the autocorrelation function and partial autocorrelation function in the energy recovery change curve, and determine the parameters p and q; The prediction model is fitted using the maximum likelihood estimation method, and the energy recovery sequence is predicted based on the prediction model to obtain the predicted sequence. By predicting the sequence, the variable relationship between energy recovery trends and electric vehicle status is evaluated, and the kinetic energy recovery strategy is optimized.

[0053] It is understood that the energy recovery data includes multiple charging power values. This invention uses the ARMA model for stationary sequence prediction, which is reflected in the predictive analysis and trend change analysis of energy recovery efficiency. The autocorrelation function (ACF) and partial autocorrelation function (PACF) are the core steps in constructing the ARMA model. Here, due to the uncertainty in the number of selected preferred time points, and the high correlation between the change characteristics of the energy recovery sequence and the energy recovery change curve, the energy recovery change curve is used as the raw data for autocorrelation function and partial autocorrelation function analysis. This improves the analysis process of the traditional ARMA model and determines the p and q parameters based on this, rather than using existing sequences for correlation function analysis. This can predict the non-stationary change characteristics of energy recovery, which is helpful for subsequent correlation assessment of energy recovery trends and vehicle state factors.

[0054] Parameter determination includes: determining the order (p) of the autoregressive component and the order (q) of the moving average component.

[0055] The electric vehicle status can include information such as braking pressure, wheel speed, deceleration, braking intensity, and braking torque. Based on the changes in vehicle status and the predicted trend of energy recovery status, the energy recovery trend of different driving modes is correlated and evaluated, and the optimal driving mode in local time is dynamically selected. In addition, by combining the analysis of vehicle braking intensity and the calculation of the maximum allowable charging power of the energy recovery process by MCU and VCU, the control period and control parameters can be optimized, thereby realizing an adaptive intelligent system that dynamically finds and maintains the optimal energy recovery point, providing effective data support for the research on kinetic energy recovery of electric vehicles.

[0056] According to an embodiment of the present invention, S7 further includes: Linear regression was performed to fit the parameter variation curves, and the goodness of fit of the parameter variation curves was analyzed. ; judge Is it greater than the preset goodness? If so, the energy recovery sequence is predicted by linear regression, with a prediction step size of N time points, to obtain the predicted sequence; Combination By comparing the predicted sequence with the variables of energy recovery trend and electric vehicle status, we can assess the relationship between these variables.

[0057] Understandably, for kinetic energy recovery test cycles with a certain degree of linear variation, linear regression can be used for fitting and sequence prediction, and the sequence prediction effect is better than that of the ARMA model.

[0058] The parameter variation curve can be evaluated using one or more parameters. Generally speaking, The goodness of fit is based on the average goodness of fit of the curves showing the variation of three parameters. Goodness of fit reflects both the linearity of kinetic energy recovery and its predictability and effectiveness. The larger the value, the greater the predictability and predictive effectiveness. The prediction step size is specifically the length formed by N consecutive time points.

[0059] A third aspect of the present invention also provides a computer-readable storage medium comprising an intelligent recovery control program for electric vehicle braking energy, wherein when executed by a processor, the intelligent recovery control program for electric vehicle braking energy implements the steps of the intelligent recovery control method for electric vehicle braking energy as described in any of the preceding claims.

[0060] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0061] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0062] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0063] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0064] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0065] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for intelligent recovery control of braking energy in electric vehicles, characterized in that, include: S1: During a test period, the target electric vehicle is powered on for testing to determine whether the vehicle is in the expected state of kinetic energy recovery. If so, the braking intensity is calculated in real time. S2: When the braking intensity is low, the VCU selects the target energy recovery level according to the driving mode, calculates the target deceleration according to the vehicle speed, performs PI control by comparing the target deceleration with the actual vehicle acceleration value issued by the current ABS, and calculates the target torque of the motor. S3: When the braking intensity is medium to high, the VCU calculates the vehicle's stability safety factor based on the braking pressure, wheel speed, and deceleration parameters, and calculates the total braking torque. It then subtracts the real-time motor torque from the total braking torque and uses the calculation results to feed back and control the hydraulic braking torque. S4: After comprehensively comparing the maximum energy recovery torque, braking torque, and maximum allowable torque, the VCU outputs the torque control result to the MCU for motor control; S5: Divide a test period into multiple time points; S6: Analyze the numerical changes of braking pressure, wheel speed, and deceleration parameters during the test period, generate parameter change curves, generate energy recovery change curves based on the charging power of the kinetic energy recovery process, introduce particle swarm optimization algorithm, use braking pressure, wheel speed, and deceleration as decision variables, set parameter constraints, use maximizing energy recovery as the objective function to search for local optimal parameters, and select the preferred time point based on the time point where the local optimal parameters are located. S7: Select the corresponding energy recovery data by optimizing the time points, serialize the energy recovery data, perform sequence prediction through the prediction model, evaluate the variable relationship between energy recovery trend and electric vehicle state, and optimize the kinetic energy recovery strategy; The maximum permissible torque is calculated as follows: Collect information on the voltage and temperature of individual battery cells in the target electric vehicle, and calculate the maximum allowable charging power of the battery; The actual battery power is calculated by collecting the actual battery pack voltage and total output current of the target electric vehicle. Calculate the charging power deviation as follows: Actual battery power - Maximum allowable battery charging power; When the charging power deviation is less than the set value of 1: Set the power rate decrease value; calculate the maximum allowable charging power = maximum allowable battery charging power - (power rate decrease * dt); When the charging power deviation is less than the set value of 2: Set the incremental power rate value; calculate the maximum allowable charging power = maximum allowable battery charging power + (incremental power rate * dt); otherwise: Maximum allowable charging power = maximum allowable charging power of the battery; Calculate the maximum regenerative braking power = maximum allowable charging power + accessory power; Calculate the maximum permissible torque for regenerative braking = maximum regenerative braking power * 9550 / current motor speed; Specifically, S6 is: During the test period, numerical changes in braking pressure, wheel speed, and deceleration parameters were analyzed, and curves showing the changes in these three parameters were generated. Based on the change analysis of charging power during the kinetic energy recovery process, an energy recovery change curve is generated; A particle swarm optimization algorithm is introduced, the number of particles M is set, and each particle corresponds to a combination of three parameters. Using braking pressure, wheel speed, and deceleration as decision variables, parameter constraints are set based on the maximum theoretical parameters of the vehicle state. In the parameter constraints, add coupling constraints, specifically using the fact that braking pressure and wheel speed are positively correlated with deceleration as constraints. In the energy recovery change curve, the goal function is to maximize the charging power. A local optimum search is performed on all particles, and the local optimum parameter combination is obtained based on the search results. In the local optimal parameter combination, the local optimal time T is marked by the parameter change curve, and the nearest time point is selected from multiple time points based on time T and marked as the preferred time point.

2. The intelligent regenerative braking control method for electric vehicles according to claim 1, characterized in that, The expected state includes: A high-voltage power-on test was conducted on the target electric vehicle. The vehicle was in the Ready state, in the D gear, and the accelerator pedal was released.

3. The intelligent regenerative braking control method for electric vehicles according to claim 1, characterized in that, In S5, for a test period, a preset time interval is set to divide multiple time points, and the interval between adjacent time points remains consistent.

4. The intelligent regenerative braking control method for electric vehicles according to claim 1, characterized in that, The S6 also includes multiple local optimal times T.

5. The intelligent regenerative braking control method for electric vehicles according to claim 1, characterized in that, Specifically, S7 is: Construct a prediction model based on ARMA; Based on the preferred time points, energy recovery data is extracted from the energy recovery change curve; The energy recovery data is time-series converted to form an energy recovery sequence; The energy recovery sequence was verified for stationarity using ADF, and the non-stationary parts were processed by differential processing. Calculate the autocorrelation function and partial autocorrelation function in the energy recovery change curve, and determine the parameters p and q; The prediction model is fitted using the maximum likelihood estimation method, and the energy recovery sequence is predicted based on the prediction model to obtain the predicted sequence. By predicting the sequence, the variable relationship between energy recovery trends and electric vehicle status is evaluated, and the kinetic energy recovery strategy is optimized.

6. The intelligent regenerative braking control method for electric vehicles according to claim 1, characterized in that, The S7 also includes: Linear regression was performed to fit the parameter variation curves, and the goodness of fit of the parameter variation curves was analyzed. ; judge Is it greater than the preset goodness? If so, the energy recovery sequence is predicted by linear regression, with a prediction step size of N time points, to obtain the predicted sequence; Combination By comparing the predicted sequence with the variables of energy recovery trend and electric vehicle status, we can assess the relationship between these variables.

7. A smart regenerative braking control system for electric vehicles, characterized in that, The system includes: a memory, a processor, and a data interface / communication interface. The memory includes an intelligent regenerative braking control program for electric vehicles. When the processor executes the intelligent regenerative braking control program for electric vehicles, it implements the steps of the intelligent regenerative braking control method for electric vehicles as described in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes an electric vehicle braking energy intelligent recovery control program, which, when executed by a processor, implements the steps of the electric vehicle braking energy intelligent recovery control method as described in any one of claims 1 to 6.

Images