Intelligent driving method and system for an automobile

Abstract

This invention provides an intelligent driving method and system for automobiles, comprising: collecting sample data of the vehicle under different road conditions; cleaning and normalizing the sample data to form training samples; inputting the training samples into a neural network for training to obtain a precise engine torque control model under different road conditions; using an onboard LiDAR on the target vehicle to send laser signals to the road surface and determining the vehicle's driving road conditions based on the laser echo signals; and using the corresponding precise engine torque control model under the road conditions to complete the torque control of the target vehicle. This invention, by sending laser signals from an onboard LiDAR to determine the vehicle's driving road conditions and simultaneously collecting and analyzing various data samples as training data to complete the training of the precise engine torque control model, can improve the vehicle's acceleration performance, steering agility, and fuel efficiency, enabling the vehicle to adapt to different road conditions and improving driving safety and stability.

Description

Technical Field

[0001] This invention relates to the field of automotive drive technology, and more specifically, to an intelligent drive method and system for automobiles. Background Technology

[0002] Currently, torque control in automotive engines is typically adjusted based on parameters such as the driver's accelerator pedal input and the engine's intake air volume. However, this method cannot fully account for the impact of factors such as road conditions and ambient temperature on the engine's torque demand, resulting in the vehicle's driving performance and fuel economy not being optimized under different road conditions. Summary of the Invention

[0003] To address the aforementioned problems, the present invention aims to provide an intelligent driving method and system for automobiles.

[0004] A smart drive method for automobiles, comprising:

[0005] Step 1: Collect sample data of the vehicle under different road conditions; the sample data includes vehicle speed, ambient temperature, coolant temperature, oil temperature, fuel consumption, road surface adhesion coefficient, driving force at the wheels, and corresponding engine torque output;

[0006] Step 2: Perform data cleaning and normalization on the sample data to form training samples;

[0007] Step 3: Input the training samples into the neural network for training to obtain a precise engine torque control model under different road conditions;

[0008] Step 4: Use the onboard LiDAR on the target vehicle to send laser signals to the road surface, and determine the road conditions based on the laser echo signals;

[0009] Step 5: Use the engine torque precision control model under the corresponding road conditions to complete the torque control of the target vehicle.

[0010] Preferably, in step 1, the fuel consumption is calculated through the following steps:

[0011] The engine's fuel consumption is determined based on the steady-state fuel consumption rate at different engine speeds; the formula for calculating engine fuel consumption is as follows:

[0012]

[0013] In the formula, Represented as the transient fuel consumption gain coefficient, A m ω represents the first regression coefficient. e Indicates engine speed, B m C represents the second regression coefficient.m D represents the third regression coefficient. m E represents the fourth regression coefficient. m Represents a constant term. T represents the engine speed change rate. e Indicates engine torque. Indicates the engine torque conversion rate. Indicates the engine's fuel consumption, b fs T represents the engine's fuel consumption rate at steady state, ρ represents fuel density, and T represents fuel density. ei Φ represents the indicated torque, and Φ represents the throttle opening.

[0014] Preferably, in step 1, the driving force at the wheel is calculated through the following steps:

[0015] Step 1.1: Determine the driving resistance based on the vehicle's condition during operation;

[0016] Step 1.2: Determine the driving force at the wheels based on the magnitude of the driving resistance; the formula for calculating the driving force at the wheels is:

[0017]

[0018] In the formula, F r F represents the driving resistance. d The driving force at the wheels is represented by m, the mass of the vehicle is represented by g, the acceleration due to gravity is represented by α, the slope angle is represented by ρ, and the air density is represented by C. D δ represents the air resistance coefficient, A represents the frontal area of ​​the vehicle, v represents the vehicle speed, δ represents the rotational mass conversion factor, and t represents time.

[0019] Preferably, step 4: using the vehicle-mounted lidar on the target vehicle to send laser signal points to the road surface, and determining the vehicle's driving conditions based on the laser echo signals, includes:

[0020] Step 4.1: Calculate the average distance between each laser signal point and a preset number of adjacent laser signal points;

[0021] Step 4.2: Calculate the distance parameters based on the average distance between each laser signal point;

[0022] Step 4.3: Determine the confidence radius of the laser signal point based on the distance parameters;

[0023] Step 4.4: Obtain the number of all laser signal points for each laser signal point within the corresponding confidence radius;

[0024] Step 4.5: When the total number of laser signal points is less than a preset value, remove the corresponding laser signal points as isolated points to obtain the laser signal points after removing isolated points;

[0025] Step 4.6: Use the laser signal points after removing isolated points to construct training samples, and input them into the neural network model for training to obtain the driving road condition detection model;

[0026] Step 4.7: Determine the current road conditions of the vehicle using the driving road condition detection model.

[0027] Preferably, step 4.2: calculating the distance parameter based on the average distance between each laser signal point includes:

[0028] Formula used:

[0029]

[0030] Calculate the distance parameters; where, This represents the distance parameter, and n represents the number of all laser signal points. d represents the average distance between the i-th laser signal point and its k neighboring laser signal points. ij This represents the Euclidean distance between the i-th laser signal point and its j-th adjacent laser signal point. This represents the sum of the average distances between all laser signal points and their corresponding k neighboring laser signal points.

[0031] Preferably, step 4.3: determining the confidence radius of the laser signal point based on the distance parameter includes:

[0032] Step 4.3.1: Calculate the standard deviation based on the Euclidean distance and average distance between adjacent laser signal points;

[0033] Step 4.3.2: Determine the confidence radius of the laser signal point using the standard deviation and distance parameters; wherein, the confidence radius is calculated using the following formula:

[0034]

[0035] Where σ is the standard deviation and r represents the confidence radius.

[0036] Preferably, step 4.6: constructing training samples using laser signal points after removing isolated points, and inputting them into a neural network model for training to obtain a driving road condition detection model, includes:

[0037] The power of the laser echo signal at each laser signal point is used as a training sample and input into the neural network model for training to obtain the driving road condition detection model.

[0038] Preferably, in step 4.6, the formula for calculating the power of the laser echo signal is:

[0039]

[0040] Among them, P r R represents the power of the laser echo signal, and D represents the detection distance. r Indicates the receiver aperture, β t η represents the laser beamwidth. atm η represents the influencing factor of laser echo signal during atmospheric transmission. sys The attenuation factor of the lidar system is represented by ρ, the reflectivity of the laser is represented by θ, the incident angle of the laser is represented by Ω, and the scattering angle of the laser is represented by A. s This represents the contact area of ​​the laser signal point.

[0041] The present invention also provides an intelligent drive system for automobiles, comprising:

[0042] The data acquisition module is used to collect sample data of the vehicle under different road conditions; the sample data includes vehicle speed, ambient temperature, coolant temperature, oil temperature, fuel consumption, road surface adhesion coefficient, driving force at the wheels, and corresponding engine torque output.

[0043] The data preprocessing module is used to clean and normalize the sample data to form training samples;

[0044] The training module is used to input the training samples into the neural network for training to obtain a precise engine torque control model under different road conditions;

[0045] The driving road condition determination module is used to send laser signals to the road surface using the vehicle-mounted LiDAR on the target vehicle, and determine the driving road condition of the vehicle based on the laser echo signal.

[0046] The torque control module is used to achieve torque control of the target vehicle using a precise engine torque control model under corresponding road conditions.

[0047] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described intelligent drive method for automobiles.

[0048] The beneficial effects of the intelligent driving method and system for automobiles provided by this invention are as follows: Compared with the prior art, this invention sends laser signals through vehicle-mounted lidar to determine the driving conditions of the vehicle, and at the same time collects and analyzes various data samples as training data to complete the training of the engine torque precision control model, which can improve the acceleration performance, steering flexibility and fuel efficiency of the vehicle, enabling the vehicle to adapt to different road conditions and improve driving safety and stability.

[0049] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 A flowchart of an intelligent driving method for automobiles provided by an embodiment of the present invention is shown;

[0052] Figure 2 The diagram shows a schematic of an intelligent drive system for automobiles provided by an embodiment of the present invention. Detailed Implementation

[0053] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0054] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0055] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0056] Please see Figure 1A smart drive method for automobiles, comprising:

[0057] Step 1: Collect sample data of the vehicle under different road conditions; the sample data includes vehicle speed, ambient temperature, coolant temperature, oil temperature, fuel consumption, road surface adhesion coefficient, driving force at the wheels, and corresponding engine torque output;

[0058] In practical applications, it is necessary to drive the test vehicle under different road conditions (rainy asphalt road, dry asphalt road, dry mud road, wet mud road, gravel road, etc.), install corresponding sensors to collect relevant sample data, and use expert analysis methods to calibrate the optimal industrial control of the vehicle under different road conditions.

[0059] In step 1 above, the fuel consumption is calculated using the following steps:

[0060] The engine's fuel consumption is determined based on the steady-state fuel consumption rate at different engine speeds; the formula for calculating engine fuel consumption is as follows:

[0061]

[0062] In the formula, Represented as the transient fuel consumption gain coefficient, A m ω represents the first regression coefficient. e Indicates engine speed, B m C represents the second regression coefficient. m D represents the third regression coefficient. m E represents the fourth regression coefficient. m Represents a constant term. T represents the engine speed change rate. e Indicates engine torque. Indicates the engine torque conversion rate. Indicates the engine's fuel consumption, b fs T represents the engine's fuel consumption rate at steady state, ρ represents fuel density, and T represents fuel density. ei Φ represents the indicated torque, and Φ represents the throttle opening.

[0063] It should be noted that the present invention can use the least squares method to perform regression fitting on the transient fuel consumption gain coefficient under different road conditions to obtain the regression coefficients for each road condition, and then calculate the fuel consumption of the car under different road conditions.

[0064] In step 1, the driving force at the wheel is calculated through the following steps:

[0065] Step 1.1: Determine the driving resistance based on the vehicle's condition during operation;

[0066] Step 1.2: Determine the driving force at the wheels based on the magnitude of the driving resistance;

[0067] Furthermore, this invention can calculate the driving force at the wheel based on Newton's laws of motion and aerodynamic principles:

[0068]

[0069] In the formula, F r F represents the driving resistance. d The driving force at the wheels is represented by m, the mass of the vehicle is represented by g, the acceleration due to gravity is represented by α, the slope angle is represented by ρ, and the air density is represented by C. D δ represents the air resistance coefficient, A represents the frontal area of ​​the vehicle, v represents the vehicle speed, δ represents the rotational mass conversion factor, and t represents time.

[0070] Step 2: Perform data cleaning and normalization on the sample data to form training samples;

[0071] This invention helps remove noise and outliers from sample data by performing data cleaning and normalization, thereby improving data quality and accuracy. At the same time, normalization can unify the data range of different features to the same scale, which helps to speed up the training of the model.

[0072] Step 3: Input the training samples into the neural network for training to obtain a precise engine torque control model under different road conditions;

[0073] Step 4: Use the onboard LiDAR on the target vehicle to send laser signals to the road surface, and determine the road conditions based on the laser echo signals;

[0074] Furthermore, step 4 includes:

[0075] Step 4.1: Calculate the average distance between each laser signal point and a preset number of adjacent laser signal points;

[0076] Step 4.2: Calculate the distance parameters based on the average distance between each laser signal point; using the formula:

[0077]

[0078] Calculate the distance parameters; where, This represents the distance parameter, and n represents the number of all laser signal points. d represents the average distance between the i-th laser signal point and its k neighboring laser signal points. ij This represents the Euclidean distance between the i-th laser signal point and its j-th adjacent laser signal point. This represents the sum of the average distances between all laser signal points and their corresponding k neighboring laser signal points;

[0079] Step 4.3: Determine the confidence radius of the laser signal point based on the distance parameters;

[0080] Furthermore, step 4.3 includes:

[0081] Step 4.3.1: Calculate the standard deviation based on the Euclidean distance and average distance between adjacent laser signal points;

[0082] Step 4.3.2: Determine the confidence radius of the laser signal point using the standard deviation and distance parameters; wherein, the confidence radius is calculated using the following formula:

[0083]

[0084] Where σ is the standard deviation and r represents the confidence radius.

[0085] Step 4.4: Obtain the number of all laser signal points for each laser signal point within the corresponding confidence radius;

[0086] Step 4.5: When the total number of laser signal points is less than a preset value, remove the corresponding laser signal points as isolated points to obtain the laser signal points after removing isolated points;

[0087] In laser point cloud analysis, it is often necessary to remove isolated point cloud data points. This is because isolated point cloud data points may be noise points caused by sensor errors, environmental interference, or other factors, which can affect subsequent point cloud processing and analysis. Removing isolated point cloud data points can improve the quality and accuracy of point cloud data, which is beneficial for subsequent point cloud processing and analysis. This invention, based on statistical principles, can identify isolated laser signal points that are significantly distant from their neighboring laser points, thus avoiding noise affecting the determination of road conditions for vehicles.

[0088] Step 4.6: Use the laser signal points after removing isolated points to construct training samples, and input them into the neural network model for training to obtain the driving road condition detection model;

[0089] Because different road surfaces have varying materials and smoothness, their reflectivity to laser light also varies. There is a direct relationship between laser reflectivity and laser power. When a laser beam strikes an object's surface, some of the laser energy is absorbed by the surface, and some is reflected. Therefore, the higher the reflectivity, the greater the reflected laser power. Conversely, the lower the reflectivity, the lower the reflected laser power. In general, there is a positive correlation between laser reflectivity and laser power; that is, the higher the reflectivity, the greater the reflected laser power.

[0090] Therefore, this invention uses the power of the laser echo signal at each laser signal point as a training sample and inputs it into a neural network model for training to obtain a road condition detection model. This model can reflect information such as the material and smoothness of the road surface, which is crucial for real-time road condition detection. The neural network model can learn this information to achieve rapid response and accurate judgment of real-time road conditions.

[0091] The formula for calculating the power of the laser echo signal is as follows:

[0092]

[0093] Among them, P r R represents the power of the laser echo signal, and D represents the detection distance. r Indicates the receiver aperture, β t η represents the laser beamwidth. atm η represents the influencing factor of laser echo signal during atmospheric transmission. sys The attenuation factor of the lidar system is represented by ρ, the reflectivity of the laser is represented by θ, the incident angle of the laser is represented by Ω, and the scattering angle of the laser is represented by A. s This represents the contact area of ​​the laser signal point.

[0094] Step 4.7: Determine the current road conditions of the vehicle using the driving road condition detection model.

[0095] Step 5: Use the engine torque precision control model under the corresponding road conditions to complete the torque control of the target vehicle.

[0096] This invention uses an onboard lidar to send laser signals and determine the road conditions of a vehicle. At the same time, it collects and analyzes various data samples as training data to train a precise engine torque control model. This can improve the vehicle's acceleration performance, steering agility, and fuel efficiency, enabling the vehicle to adapt to different road conditions and improve driving safety and stability.

[0097] Please see Figure 2 The present invention also provides an intelligent drive system for automobiles, comprising:

[0098] The data acquisition module is used to collect sample data of the vehicle under different road conditions; the sample data includes vehicle speed, ambient temperature, coolant temperature, oil temperature, fuel consumption, road surface adhesion coefficient, driving force at the wheels, and corresponding engine torque output.

[0099] The data preprocessing module is used to clean and normalize the sample data to form training samples.

[0100] The training module is used to input the training samples into the neural network for training to obtain a precise engine torque control model under different road conditions;

[0101] The driving road condition determination module is used to send laser signals to the road surface using the vehicle-mounted LiDAR on the target vehicle, and determine the driving road condition of the vehicle based on the laser echo signal.

[0102] The torque control module is used to achieve torque control of the target vehicle using a precise engine torque control model under corresponding road conditions.

[0103] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described intelligent drive method for automobiles.

[0104] Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present invention are the same as the beneficial effects of the intelligent driving method for automobiles described in the above technical solution, and will not be repeated here.

[0105] 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 variations 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. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An intelligent driving method for automobiles, characterized in that, include: Step 1: Collect sample data of the vehicle under different road conditions; the sample data includes vehicle speed, ambient temperature, coolant temperature, oil temperature, fuel consumption, road surface adhesion coefficient, driving force at the wheels, and corresponding engine torque output; Step 2: Perform data cleaning and normalization on the sample data to form training samples; Step 3: Input the training samples into the neural network for training to obtain a precise engine torque control model under different road conditions; Step 4: Use the onboard LiDAR on the target vehicle to send laser signals to the road surface, and determine the road conditions based on the laser echo signals; Step 4: Using the vehicle-mounted LiDAR on the target vehicle to send laser signal points to the road surface, and determining the vehicle's driving conditions based on the laser echo signals, including: Step 4.1: Calculate the average distance between each laser signal point and the preset k adjacent laser signal points; Step 4.2: Calculate the distance parameters based on the average distance between each laser signal point; Step 4.3: Determine the confidence radius of the laser signal point based on the distance parameters; Step 4.4: Obtain the number of all laser signal points for each laser signal point within the corresponding confidence radius; Step 4.5: When the total number of laser signal points is less than a preset value, remove the corresponding laser signal points as isolated points to obtain the laser signal points after removing isolated points; Step 4.6: Use the laser signal points after removing isolated points to construct training samples, and input them into the neural network model for training to obtain the driving road condition detection model; Step 4.7: Determine the current road conditions of the vehicle using the aforementioned road condition detection model; Step 5: Use the engine torque precision control model under the corresponding road conditions to complete the torque control of the target vehicle; Step 4.2: Calculate the distance parameters based on the average distance between each laser signal point, including: Formula used: Calculate the distance parameters; where, This represents the distance parameter, and n represents the number of all laser signal points. d represents the average distance between the i-th laser signal point and its k neighboring laser signal points. ij This represents the Euclidean distance between the i-th laser signal point and its j-th adjacent laser signal point. This represents the sum of the average distances between all laser signal points and their corresponding k neighboring laser signal points; Step 4.3: Determining the confidence radius of the laser signal point based on the distance parameter includes: Step 4.3.1: Calculate the standard deviation based on the Euclidean distance and average distance between adjacent laser signal points; Step 4.3.2: Determine the confidence radius of the laser signal point using the standard deviation and distance parameters; wherein, the confidence radius is calculated using the following formula: Where σ is the standard deviation and r represents the confidence radius.

2. The intelligent driving method for automobiles according to claim 1, characterized in that, In step 1, the driving force at the wheel is calculated through the following steps: Step 1.1: Determine the driving resistance based on the vehicle's condition during operation; Step 1.2: Determine the driving force at the wheels based on the magnitude of the driving resistance; the formula for calculating the driving force at the wheels is: In the formula, F r F represents the driving resistance. d The driving force at the wheels is represented by m, the mass of the vehicle is represented by g, the acceleration due to gravity is represented by α, the slope angle is represented by ρ, and the air density is represented by C. D δ represents the air resistance coefficient, A represents the vehicle's frontal area, v represents the vehicle speed, δ represents the rotational mass conversion factor, and t represents time.

3. The intelligent driving method for automobiles according to claim 2, characterized in that, Step 4.6: Constructing training samples using laser signal points after removing isolated points, and inputting them into the neural network model for training to obtain the driving road condition detection model, including: The power of the laser echo signal at each laser signal point is used as a training sample and input into the neural network model for training to obtain the driving road condition detection model.

4. The intelligent driving method for automobiles according to claim 3, characterized in that, In step 4.6, the formula for calculating the power of the laser echo signal is: Among them, P r R represents the power of the laser echo signal, and D represents the detection distance. r Indicates the receiver aperture, β t η represents the laser beamwidth. atm η represents the influencing factor of laser echo signal during atmospheric transmission. sys The attenuation factor of the lidar system is represented by ρ, the reflectivity of the laser is represented by θ, the incident angle of the laser is represented by Ω, and the scattering angle of the laser is represented by A. s This represents the contact area of ​​the laser signal point.

5. An intelligent drive system for automobiles, characterized in that, include: The data acquisition module is used to collect sample data of the vehicle under different road conditions; the sample data includes vehicle speed, ambient temperature, coolant temperature, oil temperature, fuel consumption, road surface adhesion coefficient, driving force at the wheels, and corresponding engine torque output. The data preprocessing module is used to clean and normalize the sample data to form training samples; The training module is used to input the training samples into the neural network for training to obtain a precise engine torque control model under different road conditions; The driving road condition determination module is used to send laser signals to the road surface using the vehicle-mounted LiDAR on the target vehicle, and determine the driving road condition of the vehicle based on the laser echo signal. The driving condition determination module includes: Step 4.1: Calculate the average distance between each laser signal point and the preset k adjacent laser signal points; Step 4.2: Calculate the distance parameters based on the average distance between each laser signal point; Step 4.3: Determine the confidence radius of the laser signal point based on the distance parameters; Step 4.4: Obtain the number of all laser signal points for each laser signal point within the corresponding confidence radius; Step 4.5: When the total number of laser signal points is less than a preset value, remove the corresponding laser signal points as isolated points to obtain the laser signal points after removing isolated points; Step 4.6: Use the laser signal points after removing isolated points to construct training samples, and input them into the neural network model for training to obtain the driving road condition detection model; Step 4.7: Determine the current road conditions of the vehicle using the aforementioned road condition detection model; The torque control module is used to complete the torque control of the target vehicle using a precise engine torque control model under the corresponding road conditions. Distance parameters are calculated based on the average distance between each laser signal point, including: Formula used: Calculate the distance parameters; where, This represents the distance parameter, and n represents the number of all laser signal points. d represents the average distance between the i-th laser signal point and its k neighboring laser signal points. ij This represents the Euclidean distance between the i-th laser signal point and its j-th adjacent laser signal point. This represents the sum of the average distances between all laser signal points and their corresponding k neighboring laser signal points; Determining the confidence radius of the laser signal point based on the distance parameter includes: The standard deviation is calculated based on the Euclidean distance and the average distance between adjacent laser signal points; The confidence radius of the laser signal point is determined using the standard deviation and distance parameters; wherein, the formula for calculating the confidence radius is: Where σ is the standard deviation and r represents the confidence radius.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps in the intelligent drive method for a car as described in any one of claims 1-4.

Images