An intelligent pure electric vehicle braking energy recovery control method

Abstract

The present invention belongs to the field of fingerprint recognition technology and provides an intelligent pure electric vehicle braking energy recovery control method, comprising: using a camera, millimeter-wave radar, and a vehicle networking module to collect data such as front images, vehicle distance and speed, traffic flow, and road congestion, and using a convolutional neural network and logic judgment rules through an on-board processor to identify the road condition type; using an ECU to record driver operation and vehicle status data, and using a cluster analysis algorithm and a K-Means algorithm to identify aggressive, moderate, and economical driving habits; constructing an MLP neural network prediction model with vehicle speed, road condition type, and driving habit type as input layers and future vehicle speed changes and braking requirements as output layers; when it is predicted that the vehicle will enter a congested section and the driver has aggressive habits, gradually reducing the kinetic energy recovery intensity in advance at specific time intervals and amplitudes; and adjusting the recovery intensity according to a preset strategy curve if the road condition suddenly changes during driving.

Description

Technical Field

[0001] The present invention belongs to the technical field of kinetic energy recovery, and in particular to an intelligent braking energy recovery control method for a pure electric vehicle. Background Art

[0002] With growing environmental awareness and the pursuit of energy sustainability, pure electric vehicles have received widespread attention and development. Braking energy regeneration, a key technology for improving energy efficiency in pure electric vehicles, converts and stores some of the kinetic energy during braking, thereby increasing the vehicle's range.

[0003] However, existing brake energy recovery systems have some problems in practical applications, the most prominent of which is the jerkiness caused by kinetic energy recovery, which not only affects driving comfort but may also have a certain negative impact on vehicle handling performance.

[0004] In addition, current braking energy recovery control methods often fail to fully consider factors such as road conditions and driving habits, resulting in difficulty in achieving optimal energy recovery efficiency and driving experience.

[0005] To this end, the present invention provides an intelligent pure electric vehicle braking energy recovery control method. Summary of the Invention

[0006] In order to make up for the deficiencies of the prior art, at least one technical problem raised in the background technology is solved.

[0007] The technical solution adopted by the present invention to solve its technical problem is:

[0008] In a first aspect, the present invention provides an intelligent pure electric vehicle braking energy recovery control method, comprising the following steps:

[0009] Step 1: Collect driving data, traffic flow and road congestion data of the vehicle during driving, and use the road condition recognition algorithm to determine the current road condition type;

[0010] Step 2: Use the ECU to continuously record driver operation data and vehicle status data, and use the data analysis module to obtain the driver's driving habits;

[0011] Step 3: Based on the determined road condition type and analyzed driving habit data, a prediction model is constructed using a neural network algorithm. The kinetic energy recovery prediction model is then trained using this data.

[0012] Step 4: To achieve a balance between driving comfort and energy recovery efficiency under different road conditions and driving habits, it is necessary to pre-set a strategy curve for how the energy recovery strength changes with road conditions and driving habits. Using a kinetic energy recovery prediction model, the energy recovery strength can be dynamically adjusted.

[0013] The specific process of dynamically adjusting the kinetic energy recovery strength is as follows:

[0014] The average speed of the vehicle is used as an indicator to measure the degree of traffic congestion. The driver's aggressiveness score s is also taken into account. The kinetic energy recovery strength F is expressed as a percentage.

[0015] A strategy curve function is obtained by fitting the experimental data.

[0016] ;

[0017] When the prediction model predicts that the vehicle is about to enter a congested road section, and predicts that the average speed v in the next 1 minute will be lower than 30km / h and the driver has aggressive driving habits, assuming the aggressiveness score s=8, the system begins to gradually reduce the kinetic energy recovery intensity in advance.

[0018] In a second aspect, the present invention provides an intelligent pure electric vehicle braking energy recovery control system, comprising the following modules:

[0019] Data acquisition module: collects driving data, traffic flow and road congestion data of vehicles during driving, and uses road condition recognition algorithm to determine the current road condition type;

[0020] Data analysis module: Use the ECU to continuously record driver operation data and vehicle status data, and obtain the driver's driving habits through the data analysis module;

[0021] Kinetic energy recovery prediction module: Based on the determined road condition type and analyzed driving habit data, a neural network algorithm is used to build a prediction model, and the kinetic energy recovery prediction model is obtained through data training;

[0022] Regeneration Force Adjustment Module: To achieve a balance between driving comfort and energy recovery efficiency under different road conditions and driving habits, it is necessary to pre-set a strategy curve for how the kinetic energy recovery force changes with road conditions and driving habits. Using a kinetic energy recovery prediction model, the kinetic energy recovery force is dynamically adjusted.

[0023] The specific process of dynamically adjusting the kinetic energy recovery strength is as follows:

[0024] The average speed of the vehicle is used as an indicator to measure the degree of traffic congestion. The driver's aggressiveness score s is also taken into account. The kinetic energy recovery strength F is expressed as a percentage.

[0025] A strategy curve function is obtained by fitting the experimental data.

[0026] ;

[0027] When the prediction model predicts that the vehicle is about to enter a congested road section, and predicts that the average speed v in the next 1 minute will be lower than 30km / h and the driver has aggressive driving habits, assuming the aggressiveness score s=8, the system begins to gradually reduce the kinetic energy recovery intensity in advance.

[0028] The beneficial effects of the present invention are as follows:

[0029] 1. It can predict in advance when the vehicle is about to enter a congested road section and the driver has aggressive driving habits, and gradually reduce the kinetic energy recovery force in advance to avoid the feeling of frustration caused by excessive recovery force during frequent starts and stops, greatly improving the riding experience during the driving process.

[0030] 2. Through in-depth analysis of driver operation data and vehicle status data, the system accurately identifies driving habits and establishes a predictive model based on road conditions. This allows the kinetic energy recovery strategy to better adapt to different drivers' driving styles, maximizing braking energy recovery and increasing vehicle range while ensuring comfort. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The present invention will be further described below with reference to the accompanying drawings.

[0032] Figure 1 This is a flowchart of the steps of an intelligent pure electric vehicle braking energy recovery control method of the present invention;

[0033] Figure 2 The present invention is a module schematic diagram of an intelligent pure electric vehicle braking energy recovery control system. DETAILED DESCRIPTION

[0034] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0035] Example 1

[0036] like Figure 1 As shown, an intelligent pure electric vehicle braking energy recovery control method according to an embodiment of the present invention includes:

[0037] Step 1: Collect driving data, traffic flow and road congestion data of the vehicle during driving, and use the road condition recognition algorithm to determine the current road condition type;

[0038] During the driving process, the camera captures the image in front of the vehicle in real time, including traffic signs and lane lines; the millimeter-wave radar monitors the distance and speed of the vehicle in front in real time; and the Internet of Vehicles module obtains the latest traffic flow and road congestion data from the cloud server.

[0039] The collected information is transmitted to the on-board processor, and the current road condition type is determined through road condition recognition algorithm analysis;

[0040] For example, if it is detected that the distance between the vehicles ahead is small and the speed changes frequently, it is determined to be a congested road section in combination with traffic flow data;

[0041] Specifically, the millimeter-wave radar can monitor the speed information of the vehicle in front in real time; it calculates the speed of the vehicle in front by the Doppler frequency shift method, and the formula is: ,in, is the Doppler shift, which is the frequency difference between the transmitted and reflected waves, is the radial velocity of the target object relative to the radar, is the wavelength of the millimeter wave, is the frequency of the radar's transmitted signal;

[0042] Obtained by measurement , the transmission frequency of the millimeter wave radar is known and wavelength , the relative speed of the target object can be calculated ;

[0043] Millimeter-wave radar can monitor the distance information of the vehicle in front in real time and calculate the distance of the vehicle in front based on the time delay principle. The formula is: , R is the distance between the target object and the radar, c is the speed of light, is the time delay between the transmitted and received signals;

[0044] Since the time delay between the transmission and reception of millimeter-wave radar signals is proportional to the target distance, by measuring the precise , the distance R between the target object and the vehicle can be calculated;

[0045] Based on real-time front-view images captured by the camera, millimeter-wave radar monitors the distance and speed of the vehicle ahead, and the Internet of Vehicles module obtains traffic flow and road congestion data from the cloud;

[0046] The collected driving data, traffic flow and road congestion data are transmitted to the on-board processor, and the road condition recognition algorithm is used to determine the road condition type;

[0047] Specifically, a convolutional neural network (CNN) algorithm is used to process camera images to identify traffic signs, lane lines, and the distance and speed of the vehicle ahead, and a road condition recognition model is trained.

[0048] Based on the road condition recognition model, feature extraction and classification are performed. When judging the road condition based on the changes in vehicle spacing and speed, simple logical judgment rules are applied. For example, if the vehicle spacing is set to be less than a certain threshold and the speed standard deviation is greater than a certain value, combined with traffic flow data, if the number of vehicles passing through a certain section per unit time exceeds a set value, the section is judged to be congested or non-congested.

[0049] Step 2: Use the ECU to continuously record driver operation data and vehicle status data, and use the data analysis module to obtain the driver's driving habits;

[0050] The driving habit recognition and analysis step uses the ECU to continuously record driver operation data and vehicle status data, and regularly transmits the data to the data analysis module; the data analysis module uses machine learning algorithms to process and analyze the data to obtain the driver's driving habits;

[0051] For example, by analyzing the frequency and amplitude of changes in the opening of the accelerator pedal and the brake pedal, it is determined whether the driver has an aggressive, moderate, or economical driving habit;

[0052] Specifically, the ECU continuously records driver operation data, including accelerator pedal opening, brake pedal opening, gear shifting, steering angle, and vehicle status data, including vehicle speed, acceleration, and deceleration, and regularly transmits them to the data analysis module;

[0053] Cluster analysis algorithm is used to standardize the operation data and vehicle status data, and the formula is: , where x is the original data, μ is the data mean, and σ is the standard deviation. The data is normalized; the operation data of different drivers are mapped to the same scale; then, the K-Means algorithm is used to cluster the data into different categories corresponding to aggressive, moderate, and economical driving habits;

[0054] Specifically, for example, for driver A, during a certain period of driving, the accelerator pedal opening data is [10%, 12%, 15%, 13%, 11%, 10%, 14%, 16%, 12%, 13%], the brake pedal opening data is [0%, 0%, 0%, 5%, 0%, 0%, 0%, 0%, 0%, 3%], and the vehicle speed data is [30km / h, 32km / h, 35km / h, 33km / h, 31km / h, 30km / h, 34km / h, 36km / h, 32km / h, 33km / h];

[0055] At the same time, the gear shift operation and steering angle are recorded. For example, during a turn, the steering angle gradually increases from 0° to 30° and then gradually returns to the normal value. The acceleration and deceleration are calculated based on the vehicle speed data.

[0056] The ECU packages all the data recorded during this period and transmits it to the data analysis module via the in-vehicle network. The analysis module then analyzes the data. For example, let's assume the ECU records the mean μ and standard deviation σ of all accelerator pedal opening data for 10 drivers over a week. Assuming the mean μ = 15% and the standard deviation σ = 3%, the module calculates the mean μ = 15% and the standard deviation σ = 3%.

[0057] For driver A's first accelerator pedal opening data of 10%, the normalized value is All accelerator pedal opening data of driver A and the corresponding data of the other nine drivers were normalized in this way so that all drivers' data were on the same scale, which facilitated subsequent analysis.

[0058] Determine the number of clusters and classify driving habits into three categories: aggressive, moderate, and economical, so K=3. Initialize the cluster centers: Randomly select three points from all standardized accelerator pedal opening and brake pedal opening data points as the initial cluster centers c1, c2, and c3. Assume that the initial cluster center c1 corresponds to a standardized accelerator pedal opening value of -0.5 and a standardized brake pedal opening value of 0.2; c2 corresponds to a standardized accelerator pedal opening value of 1.2 and a standardized brake pedal opening value of -0.8; c3 corresponds to a standardized accelerator pedal opening value of 0.3 and a standardized brake pedal opening value of 0.1.

[0059] Calculate the distance from the data point to the cluster center. For the first data point of driver A, that is, the normalized value of the accelerator pedal opening is -1.67 and the normalized value of the brake pedal opening is 0, calculate its distance to the three cluster centers.

[0060] Distance to c1 ;

[0061] Distance to c2 ;

[0062] Distance to c3 ;

[0063] Since the distance to c1 is the shortest, the data point is classified as the category of c1;

[0064] After traversing all data points of all drivers and completing classification, recalculate the cluster center of each category. For example, for category c1, assuming there are 2000 data points in this category, recalculate the average of the standardized values ​​of the accelerator pedal opening and the brake pedal opening to obtain the new cluster center.

[0065] Repeat the above steps of calculating distance and updating cluster centers until the cluster centers no longer change significantly;

[0066] If the normalized values ​​of the accelerator pedal opening of the data points in a cluster are generally large and fluctuate frequently, and the normalized values ​​of the brake pedal opening are also large and appear frequently, then the driving habit corresponding to this cluster is aggressive;

[0067] If the normalized values ​​of the accelerator pedal opening of the data points in a cluster are relatively stable and small, and the normalized values ​​of the brake pedal opening are also small and occur less frequently, it corresponds to a mild driving habit;

[0068] For economical driving habits, the accelerator pedal opening is moderate and changes smoothly, and the brake pedal opening is used less frequently, which corresponds to the characteristics of the cluster where the cluster center c3 is located;

[0069] Step 3: Based on the determined road condition type and analyzed driving habit data, a prediction model is constructed using a neural network algorithm. The kinetic energy recovery prediction model is then trained using this data.

[0070] A prediction model is established based on collected road condition information and driving habit data. A neural network algorithm is used to construct the model, with vehicle speed, road condition type, and driving habit type as input layers, and future speed changes and braking requirements as output layers. A kinetic energy recovery prediction model is developed through data training. As the vehicle is driving, new actual driving data is continuously input into the kinetic energy recovery prediction model. Model parameters are adjusted through optimization algorithms such as backpropagation algorithms to make the kinetic energy recovery prediction model more accurate.

[0071] Specifically, the system uses vehicle speed, road condition type, driving habits, etc. as input layers, and vehicle speed changes and braking demands as outputs, where the expected brake pedal opening percentage is expressed. An MLP neural network with one hidden layer containing five neurons is constructed, and a weighted sum calculation is performed on the hidden layer.

[0072] For example, for the first neuron j=1 in the hidden layer, the input layer has three inputs: x1 corresponds to the vehicle speed, x2 corresponds to the road condition type, i.e., congested road section or non-congested road section, and x3 corresponds to the driving habit type; the corresponding output weights are w11, w12, w13; and the bias term b1;

[0073] It should be noted that, for the convenience of calculation and training, the congested road section in the road condition type is marked as 1, and the non-congested road section is marked as 0; in the driving habit type, the aggressive type is set to 0, the moderate type is set to 1, and the economical type is set to 2;

[0074] For example, assuming w11=0.3, w12=−0.1, w13=0.2, b1=0.05, for a set of input data x1=50, i.e., a vehicle speed of 50 km / h, x2=1, i.e., a congested road section, x3=1, i.e., a mild driving habit, the input z1 of the neuron is: ; where i=1,2,3;

[0075] Similarly, for other neurons in the hidden layer, weighted summation is calculated according to this formula;

[0076] The ReLU activation function is used, f(z)=max(0,z); for z1=15.15, after being processed by the ReLU activation function, f(z1)=max(0,15.15)=15.15; if the value of z is negative, such as z=−2, then f(z)=max(0,−2)=0; the output of each neuron in the hidden layer is processed by the ReLU activation function and used as the input of the next layer (output layer);

[0077] Assume that the output layer has two neurons, corresponding to the speed change and braking demand in the next 10 seconds respectively; there are also weights and biases between the output layer neurons and the hidden layer neurons;

[0078] For example, the weights of the first neuron in the output layer, which corresponds to the prediction of vehicle speed change, and the weights of the five neurons in the hidden layer are v11, v12, v13, v14, and v15, respectively. The bias term is e1. Assuming v11=0.2, v12=0.1, v13=−0.05, v14=0.15, v15=0.08, and e1=−0.1, the outputs of the five neurons in the hidden layer after ReLU activation are h1=15.15, h2=10.2, h3=0, h4=8.5, and h5=12.3, respectively. Then the output of the first neuron in the output layer is: ; where j = 1, 2, 3, 4, 5; similarly, the output of the second neuron in the output layer can be calculated;

[0079] The mean square error loss function is used to optimize the model, and the formula is: , where m is the number of training data sets. Assume that for a certain set of data, the actual speed change in the next 10 seconds is , predicted value , the actual braking demand , predicted value , then the loss of this group of data is 0.0312625: sum up the losses of all m groups of data and take the average to get the loss L of the entire training set;

[0080] The backpropagation algorithm is then used to update the parameters. Based on the principle of gradient descent, the gradient of the loss function with respect to the weights and biases is calculated. After the gradient is calculated, the weights and biases are updated according to the gradient descent formula. The above process of calculating the loss function and updating the parameters is repeated. After multiple rounds of iteration, the loss function is gradually reduced, and the prediction accuracy of the kinetic energy recovery prediction model is continuously improved.

[0081] Step 4: Utilize the kinetic energy recovery prediction model to dynamically adjust the kinetic energy recovery intensity based on a pre-set strategy curve that shows how the kinetic energy recovery intensity changes with road conditions and driving habits;

[0082] Based on the kinetic energy recovery prediction model, when the model predicts that the vehicle is about to enter a congested section and the driver has aggressive driving habits, the kinetic energy recovery force is gradually reduced in advance to avoid vehicle jerking due to excessive regeneration force during frequent starts and stops;

[0083] If the road condition suddenly changes from smooth to congested during driving, the system will adjust the kinetic energy recovery strategy in real time, reducing the recovery force according to the preset strategy curve to ensure a balance between driving comfort and energy recovery efficiency.

[0084] Specifically, in order to achieve a balance between driving comfort and energy recovery efficiency under different road conditions and driving habits, it is necessary to pre-set a strategy curve that changes the kinetic energy recovery intensity with road conditions and driving habits;

[0085] For example, let's assume we use the average vehicle speed as an indicator of traffic congestion, while also considering the driver's aggressiveness score s (a higher score indicates more aggressive driving habits, ranging from 0 to 10); and the kinetic energy recovery force F is expressed as a percentage (ranging from 0 to 100%).

[0086] A strategy curve function is obtained by fitting the experimental data, such as ;

[0087] When the prediction model predicts that the vehicle is about to enter a congested section, for example, if the average speed v is predicted to be less than 30 km / h in the next minute and the driver has aggressive driving habits (assuming the aggressiveness score s=8), the system will begin to gradually reduce the kinetic energy recovery force in advance;

[0088] Assume that the current vehicle's kinetic energy recovery strength is , current vehicle speed Based on the predicted traffic congestion and driving habits, we can calculate the target kinetic energy recovery strength. ;

[0089] In order to avoid setbacks caused by sudden changes in the recovery force, the system Gradually reduce the kinetic energy recovery strength; the magnitude of each reduction It can be based on the remaining time, such as the predicted time to enter the congested section and the target reduction range;

[0090] If the vehicle suddenly changes from smooth traffic to congested traffic during driving, the system needs to adjust the kinetic energy recovery strategy in real time. The system immediately calculates the required adjustment range based on the current kinetic energy recovery strength and the new target kinetic energy recovery strength. Similarly, to avoid setbacks, the system reduces the kinetic energy recovery strength at a certain adjustment rate.

[0091] The technical solution of this embodiment is as follows: by using cameras, millimeter-wave radar, and a vehicle networking module to collect data such as forward images, vehicle distance and speed, traffic flow, and road congestion, the on-board processor uses a convolutional neural network and logical judgment rules to identify road condition types; the ECU records driver operation and vehicle status data, and uses cluster analysis algorithms and K-Means algorithms to identify aggressive, moderate, and economical driving habits; an MLP neural network prediction model is constructed using vehicle speed, road condition type, and driving habit type as the input layer, and future vehicle speed changes and braking requirements as the output layer, and is optimized using a backpropagation algorithm to make predictions more accurate; when it is predicted that the vehicle will enter a congested section and the driver has aggressive habits, the kinetic energy recovery force is gradually reduced in advance at specific time intervals and amplitudes; when the road condition suddenly changes during driving, the recovery force is adjusted in real time according to a preset strategy curve to ensure a balance between driving comfort and energy recovery efficiency, thereby improving the overall performance of the pure electric vehicle.

[0092] Example 2

[0093] like Figure 2 Based on Example 1, the present invention further provides an intelligent pure electric vehicle braking energy recovery control system, including the following modules:

[0094] Data acquisition module: collects driving data, traffic flow and road congestion data of vehicles during driving, and uses road condition recognition algorithm to determine the current road condition type;

[0095] During the driving process, the camera captures the image in front of the vehicle in real time, including traffic signs and lane lines; the millimeter-wave radar monitors the distance and speed of the vehicle in front in real time; and the Internet of Vehicles module obtains the latest traffic flow and road congestion data from the cloud server.

[0096] The collected information is transmitted to the on-board processor, and the current road condition type is determined through road condition recognition algorithm analysis;

[0097] Based on real-time front-view images captured by the camera, millimeter-wave radar monitors the distance and speed of the vehicle ahead, and the Internet of Vehicles module obtains traffic flow and road congestion data from the cloud;

[0098] The collected driving data, traffic flow and road congestion data are transmitted to the on-board processor, and the road condition recognition algorithm is used to determine the road condition type;

[0099] Specifically, a convolutional neural network (CNN) algorithm is used to process camera images to identify traffic signs, lane lines, and the distance and speed of the vehicle ahead, and a road condition recognition model is trained.

[0100] Based on the road condition recognition model, feature extraction and classification are performed. When judging the road condition based on the changes in vehicle spacing and speed, simple logical judgment rules are applied. For example, if the vehicle spacing is set to be less than a certain threshold and the speed standard deviation is greater than a certain value, combined with traffic flow data, if the number of vehicles passing through a certain section per unit time exceeds a set value, the section is judged to be congested or non-congested.

[0101] Data analysis module: Use the ECU to continuously record driver operation data and vehicle status data, and obtain the driver's driving habits through the data analysis module;

[0102] The driving habit recognition and analysis step uses the ECU to continuously record driver operation data and vehicle status data, and regularly transmits the data to the data analysis module; the data analysis module uses machine learning algorithms to process and analyze the data to obtain the driver's driving habits;

[0103] For example, by analyzing the frequency and amplitude of changes in the opening of the accelerator and brake pedals, it can be determined whether the driver has aggressive, moderate, or economical driving habits;

[0104] Specifically, the ECU continuously records driver operation data, including accelerator pedal opening, brake pedal opening, gear shifting, steering angle, and vehicle status data, including vehicle speed, acceleration, and deceleration, and regularly transmits them to the data analysis module;

[0105] Cluster analysis algorithm is used to standardize the operation data and vehicle status data, and the formula is: , where x is the original data, μ is the data mean, and σ is the standard deviation. The data is normalized; the operation data of different drivers are mapped to the same scale; then, the K-Means algorithm is used to cluster the data into different categories corresponding to aggressive, moderate, and economical driving habits;

[0106] Specifically, for example, for driver A, during a certain period of driving, the accelerator pedal opening data is [10%, 12%, 15%, 13%, 11%, 10%, 14%, 16%, 12%, 13%], the brake pedal opening data is [0%, 0%, 0%, 5%, 0%, 0%, 0%, 0%, 0%, 3%], and the vehicle speed data is [30km / h, 32km / h, 35km / h, 33km / h, 31km / h, 30km / h, 34km / h, 36km / h, 32km / h, 33km / h];

[0107] At the same time, the gear shift operation and steering angle are recorded. For example, during a turn, the steering angle gradually increases from 0° to 30° and then gradually returns to the normal value. The acceleration and deceleration are calculated based on the vehicle speed data.

[0108] The ECU packages all the data recorded during this period and transmits it to the data analysis module via the in-vehicle network. The analysis module then analyzes the data. For example, let's assume the ECU records the mean μ and standard deviation σ of all accelerator pedal opening data for 10 drivers over a week. Assuming the mean μ = 15% and the standard deviation σ = 3%, the module calculates the mean μ = 15% and the standard deviation σ = 3%.

[0109] For driver A's first accelerator pedal opening data of 10%, the normalized value is All accelerator pedal opening data of driver A and the corresponding data of the other nine drivers were normalized in this way so that all drivers' data were on the same scale, which facilitated subsequent analysis.

[0110] Determine the number of clusters and classify driving habits into three categories: aggressive, moderate, and economical, so K=3. Initialize the cluster centers: Randomly select three points from all standardized accelerator pedal opening and brake pedal opening data points as the initial cluster centers c1, c2, and c3. Assume that the initial cluster center c1 corresponds to a standardized accelerator pedal opening value of -0.5 and a standardized brake pedal opening value of 0.2; c2 corresponds to a standardized accelerator pedal opening value of 1.2 and a standardized brake pedal opening value of -0.8; c3 corresponds to a standardized accelerator pedal opening value of 0.3 and a standardized brake pedal opening value of 0.1.

[0111] Calculate the distance from the data point to the cluster center. For the first data point of driver A, that is, the normalized value of the accelerator pedal opening is -1.67 and the normalized value of the brake pedal opening is 0, calculate its distance to the three cluster centers.

[0112] Distance to c1 ;

[0113] Distance to c2 ;

[0114] Distance to c3 ;

[0115] Since the distance to c1 is the shortest, the data point is classified as the category of c1;

[0116] After traversing all data points of all drivers and completing classification, recalculate the cluster center of each category. For example, for category c1, assuming there are 2000 data points in this category, recalculate the average of the standardized values ​​of the accelerator pedal opening and the brake pedal opening to obtain the new cluster center.

[0117] Repeat the above steps of calculating distance and updating cluster centers until the cluster centers no longer change significantly;

[0118] If the normalized values ​​of the accelerator pedal opening of the data points in a cluster are generally large and fluctuate frequently, and the normalized values ​​of the brake pedal opening are also large and appear frequently, then the driving habit corresponding to this cluster is aggressive;

[0119] If the normalized values ​​of the accelerator pedal opening of the data points in a cluster are relatively stable and small, and the normalized values ​​of the brake pedal opening are also small and occur less frequently, it corresponds to a mild driving habit;

[0120] For economical driving habits, the accelerator pedal opening is moderate and changes smoothly, and the brake pedal opening is used less frequently, which corresponds to the characteristics of the cluster where the cluster center c3 is located;

[0121] Kinetic energy recovery prediction module: Based on the determined road condition type and analyzed driving habit data, a neural network algorithm is used to build a prediction model, and the kinetic energy recovery prediction model is obtained through data training;

[0122] A prediction model is established based on collected road condition information and driving habit data. A neural network algorithm is used to construct the model, with vehicle speed, road condition type, and driving habit type as input layers, and future speed changes and braking requirements as output layers. A kinetic energy recovery prediction model is developed through data training. As the vehicle is driving, new actual driving data is continuously input into the kinetic energy recovery prediction model. Model parameters are adjusted through optimization algorithms such as backpropagation algorithms to make the kinetic energy recovery prediction model more accurate.

[0123] Regeneration force adjustment module: Utilizes a kinetic energy recovery prediction model to dynamically adjust the kinetic energy recovery force based on a pre-set strategy curve that shows how the kinetic energy recovery force changes with road conditions and driving habits;

[0124] Based on the kinetic energy recovery prediction model, when the model predicts that the vehicle is about to enter a congested section and the driver has aggressive driving habits, the kinetic energy recovery force is gradually reduced in advance to avoid vehicle jerking due to excessive regeneration force during frequent starts and stops;

[0125] If the road condition suddenly changes from smooth to congested during driving, the system will adjust the kinetic energy recovery strategy in real time, reducing the recovery force according to the preset strategy curve to ensure a balance between driving comfort and energy recovery efficiency.

[0126] Specifically, in order to achieve a balance between driving comfort and energy recovery efficiency under different road conditions and driving habits, it is necessary to pre-set a strategy curve that changes the kinetic energy recovery intensity with road conditions and driving habits;

[0127] For example, let's assume we use the average vehicle speed as an indicator of traffic congestion, while also considering the driver's aggressiveness score s (a higher score indicates more aggressive driving habits, ranging from 0 to 10); and the kinetic energy recovery force F is expressed as a percentage (ranging from 0 to 100%).

[0128] A strategy curve function is obtained by fitting the experimental data, such as ;

[0129] When the prediction model predicts that the vehicle is about to enter a congested section, for example, if the average speed v is predicted to be less than 30 km / h in the next minute and the driver has aggressive driving habits (assuming the aggressiveness score s=8), the system will begin to gradually reduce the kinetic energy recovery force in advance;

[0130] Assume that the current vehicle's kinetic energy recovery strength is , current vehicle speed Based on the predicted traffic congestion and driving habits, we can calculate the target kinetic energy recovery strength. ;

[0131] In order to avoid setbacks caused by sudden changes in the recovery force, the system Gradually reduce the kinetic energy recovery strength; the magnitude of each reduction It can be based on the remaining time, such as the predicted time to enter the congested section and the target reduction range;

[0132] During vehicle driving, if the road condition suddenly changes from smooth to congested, the system needs to adjust the kinetic energy recovery strategy in real time. The system immediately calculates the adjustment range based on the current kinetic energy recovery strength and the new target kinetic energy recovery strength. Similarly, in order to avoid setbacks, the system reduces the kinetic energy recovery strength at a certain adjustment rate.

[0133] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent pure electric vehicle braking energy recovery control method, characterized by: include: Step 1: Collect driving data, traffic flow and road congestion data of the vehicle during driving, and use the road condition recognition algorithm to determine the current road condition type; Step 2: Use the ECU to continuously record driver operation data and vehicle status data, and use the data analysis module to obtain the driver's driving habits; Step 3: Based on the determined road condition type and analyzed driving habit data, a prediction model is constructed using a neural network algorithm. The kinetic energy recovery prediction model is then trained using this data. Step 4: To achieve a balance between driving comfort and energy recovery efficiency under different road conditions and driving habits, it is necessary to pre-set a strategy curve for how the energy recovery strength changes with road conditions and driving habits. Using a kinetic energy recovery prediction model, the energy recovery strength can be dynamically adjusted. The specific process of dynamically adjusting the kinetic energy recovery strength is as follows: The average speed of vehicles is used as an indicator to measure the degree of traffic congestion. The driver's aggressiveness score s is also taken into account. A higher score indicates more aggressive driving habits. The range is 0-10. The kinetic energy recovery strength F is expressed as a percentage. A strategy curve function is obtained by fitting the experimental data. ； If the prediction model predicts that the vehicle is about to enter a congested section, the average speed v will be less than 30 km / h in the next minute, and the driver has aggressive driving habits (assuming the aggressiveness score s=8), the system will begin to gradually reduce the kinetic energy recovery force in advance; Assume that the current vehicle's kinetic energy recovery strength is , current vehicle speed ; Get the target kinetic energy recovery strength based on the predicted congested road conditions and driving habits ; To avoid setbacks caused by sudden changes in recovery force, the system Gradually reduce the kinetic energy recovery strength; During vehicle driving, if the road condition suddenly changes from smooth to congested, the system needs to adjust the kinetic energy recovery strategy in real time. The system immediately calculates the required adjustment range based on the current kinetic energy recovery strength and the new target kinetic energy recovery strength. Similarly, in order to avoid setbacks, the system reduces the kinetic energy recovery strength according to the adjustment rate.

2. The intelligent pure electric vehicle braking energy recovery control method according to claim 1, characterized in that: The specific process of using the road condition recognition algorithm to determine the current road condition type is as follows: Use convolutional neural network algorithms to process camera images to identify traffic signs, lane lines, and the distance and speed of the vehicle ahead, and train a road condition recognition model; Feature extraction and classification are performed based on the road condition recognition algorithm model to determine whether the road is congested or non-congested.

3. The intelligent pure electric vehicle braking energy recovery control method according to claim 1, characterized in that: The recording of driver operation data specifically includes: Accelerator pedal opening, brake pedal opening, gear shifting operation, steering angle.

4. The intelligent pure electric vehicle braking energy recovery control method according to claim 1, characterized in that: The recorded vehicle status data includes: Vehicle speed, acceleration, and deceleration.

5. The intelligent pure electric vehicle braking energy recovery control method according to claim 1, characterized in that: The specific process of obtaining the driver's driving habits by using the data analysis module is as follows: The data analysis module uses machine learning algorithms to analyze the frequency and amplitude of changes in the opening of the accelerator and brake pedals to determine whether the driver has aggressive, moderate and economical driving habits.

6. The intelligent pure electric vehicle braking energy recovery control method according to claim 1, characterized in that: The specific process of using the neural network algorithm to build a prediction model is as follows: A neural network algorithm is used to build a model, with vehicle speed, road condition type, and driving habit type as the input layer, and vehicle speed changes, braking requirements, etc. as the output layer. A kinetic energy recovery prediction model is obtained through data training.

7. The intelligent pure electric vehicle braking energy recovery control method according to claim 1, characterized in that: The specific process of dynamically adjusting the kinetic energy recovery strength is as follows: During vehicle driving, if the road condition suddenly changes from smooth to congested, the system will adjust the kinetic energy recovery strategy in real time and reduce the recovery intensity according to the preset strategy curve.

Images