A machine learning-based energy recovery method for electric vehicles

Abstract

The present invention relates to an electric vehicle energy recovery method based on machine learning, and relates to the field of automobile driving. The method simulates the driver's daily driving habits through a machine learning algorithm to obtain a preset braking logic, and when the vehicle accelerator pedal opening and closing degree is 0, starts energy recovery according to the preset braking logic. Since the preset braking logic is obtained based on the driver's driving habits, the entire braking process is as close as possible to the driver's driving braking, so that the entire braking process conforms to the driver's habits, is more comfortable, and can recover as much energy as possible.

Description

Technical Field

[0001] The present invention relates to the field of automobile driving, and more specifically to an electric vehicle energy recovery method based on machine learning. Background Art

[0002] The existing braking energy recovery strategy in vehicle driving is set when the vehicle leaves the factory. It aims to improve the energy recovery efficiency as much as possible while ensuring driving smoothness, that is, to increase the proportion of regenerative braking force in the total braking force. However, due to different driving habits of different drivers, for example, some drivers prefer sudden braking, while some drivers prefer slow braking, resulting in a large difference in energy recovery effect. In other words, the current braking energy recovery technology in vehicle driving is optimized from the perspective of the vehicle.

[0003] For example, the existing technology uses energy recovery with a fixed intensity. When the vehicle speed is lower than a certain condition, the energy recovery will be exited. The energy recovery capacity is limited and the intensity of energy recovery is low, otherwise it will affect the driving experience and fail to fully recover the vehicle's kinetic energy. In addition, a decoupled electronic braking system is used. The intensity of motor energy recovery is determined according to the brake pedal opening. Basic braking is activated when the pedal displacement exceeds a certain value. The energy recovery intensity is a fixed value set at the factory, which may not conform to the operating habits of every driver. Summary of the invention

[0004] The purpose of the present invention is to overcome the above-mentioned shortcomings of the prior art. This patent proposes an electric vehicle energy recovery method based on machine learning, which can realize vehicle energy recovery according to the driver's driving habits.

[0005] The technical solution of the present invention is as follows:

[0006] A method for electric vehicle energy recovery based on machine learning, comprising the following steps:

[0007] Step 1, monitoring whether the opening and closing degree of the accelerator pedal of the electric vehicle is 0;

[0008] Step 2: when the opening and closing degree of the accelerator pedal is monitored to be 0, obtaining vehicle driving information and driving environment information;

[0009] Step 3, calculating the proximity items between the current vehicle driving information and the driving environment information and the preset braking logic;

[0010] Step 4: Perform energy recovery braking using the adjacent braking logic;

[0011] Step 5, determining whether the current brake pedal opening and closing degree is 0, if it is 0, braking is performed until the vehicle stops according to the adjacent braking logic, and if it is not 0, braking is performed to decelerate according to the driver's operation;

[0012] It is characterized in that the preset braking logic is obtained by learning the driver's driving habits in different vehicle driving states and driving environments through a machine learning method, and the braking logic of braking the vehicle from the current speed to 0 under different vehicle driving information and different driving environment information is obtained.

[0013] Furthermore, the vehicle driving information includes one or more of the vehicle gear position, current vehicle speed, and steering wheel rotation angle.

[0014] Furthermore, the driving environment information includes one or more of whether there is a vehicle ahead and the distance, traffic light signals, the distance to the intersection ahead, whether there is a zebra crossing ahead, whether there is an obstacle ahead and the distance to the obstacle, and weather conditions.

[0015] Furthermore, in step three, a KNN algorithm, a regression algorithm, a random forest algorithm, or a SVM algorithm is used to calculate the proximity items between the current vehicle driving information and the driving environment information and the preset braking logic.

[0016] Furthermore, the process of the driver manipulating the braking and deceleration in step five is used to correct the preset braking logic.

[0017] Furthermore, the preset braking logic is a braking logic for braking the vehicle from a current speed to a specific speed. After stopping the energy recovery braking, the vehicle prompts the driver that the energy recovery has ended.

[0018] Furthermore, the vehicle prompts the driver that within 2 seconds after the energy recovery ends, if the accelerator pedal opening and closing degrees and the brake pedal opening and closing degrees are still 0, energy recovery braking will be performed according to the adjacent braking logic until the vehicle speed is braked to 0.

[0019] Furthermore, the vehicle prompts the driver that within 2 seconds after the energy recovery ends, if the accelerator pedal opening and closing degree and the brake pedal opening and closing degree are not 0, it is controlled by the driver.

[0020] Application of any of the aforementioned machine learning-based electric vehicle energy recovery methods in reversing.

[0021] The technical solution of the present invention has the following advantages:

[0022] 1. The present invention can preset the braking logic under different vehicle driving information and driving environment information according to the driver's driving habits, so that the vehicle can recover energy according to the preset braking logic during automatic sliding, thereby maximizing energy recovery and extending the vehicle's cruising range;

[0023] 2. The present invention realizes more comfortable energy recovery, avoids the driver's aversion to the rigid logic of ordinary energy recovery, reduces the driver's fatigue, and makes the driver more comfortable. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0025] Figure 1 It is a flow chart of the electric vehicle energy recovery method based on machine learning. DETAILED DESCRIPTION

[0026] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] Example 1

[0028] Existing new energy vehicles are equipped with sensors such as laser radar and cameras, which can obtain vehicle driving information and driving environment information in real time, such as the vehicle's current speed, obstacles ahead, traffic light signals and other information. This embodiment is based on monitoring the vehicle driving information and driving environment information of the driver during driving, and continuously optimizing the driver's driving habits through a machine learning algorithm, and deriving the entire braking process from the driver applying the brakes to the vehicle stopping under different vehicle driving information and driving environment information, that is, the preset braking logic. In order to achieve maximum energy recovery during vehicle braking and make the driver feel as similar as possible to when he is driving himself, when the vehicle needs to brake and stop, a method for electric vehicle energy recovery based on machine learning is adopted. The recovery method is as shown in the attached Figure 1 As shown, it includes the following steps:

[0029] Step 1, monitoring whether the opening and closing degree of the accelerator pedal of the electric vehicle is 0;

[0030] Step 2, when the opening and closing degree of the accelerator pedal is monitored to be 0, obtaining vehicle driving information and driving environment information, such as vehicle gear position, current vehicle speed, steering wheel rotation angle, whether there is a vehicle ahead and the distance, traffic light signal, distance to the intersection ahead, whether there is a zebra crossing ahead, whether there is an obstacle ahead and the distance of the obstacle, weather conditions and other information;

[0031] Step 3, using a KNN algorithm or a regression algorithm or a random forest algorithm or a SVM algorithm to calculate adjacent items between the current vehicle driving information and the driving environment information and the preset braking logic;

[0032] Step 4: Perform energy recovery braking using the adjacent braking logic;

[0033] Step five, determine whether the current brake pedal opening and closing degree is 0. If it is 0, brake until stop according to the adjacent braking logic. If it is not 0, brake and decelerate according to the driver's operation.

[0034] The process of the driver manipulating the braking and deceleration in step 5 is used to correct the preset braking logic.

[0035] Example 2

[0036] During vehicle driving, not every braking process is braking until the vehicle stops. For example, when the vehicle ahead is decelerating, the driver only needs to brake until the vehicle speed is lower than or equal to the speed of the vehicle ahead. In order to recover the energy in this process, the preset braking logic in the electric vehicle energy recovery method based on machine learning needs to be set to the braking logic that brakes the vehicle from the current speed to a specific speed. Figure 1 As shown, it includes the following steps:

[0037] Step 1, monitoring whether the opening and closing degree of the accelerator pedal of the electric vehicle is 0;

[0038] Step 2, when the opening and closing degree of the accelerator pedal is monitored to be 0, the vehicle driving information and driving environment information are obtained, wherein the vehicle driving environment must include whether there is a vehicle ahead and the distance, and other information includes one or more of the vehicle gear position, current vehicle speed, steering wheel rotation angle, traffic light signal, distance to the intersection ahead, whether there is a zebra crossing ahead, whether there is an obstacle ahead and the distance of the obstacle, weather conditions, etc.;

[0039] Step 3, using a KNN algorithm or a regression algorithm or a random forest algorithm or a SVM algorithm to calculate adjacent items between the current vehicle driving information and the driving environment information and the preset braking logic;

[0040] Step 4: Perform energy recovery braking using the adjacent braking logic;

[0041] Step five, determine whether the current brake pedal opening and closing degree is 0. If it is 0, brake until stop according to the adjacent braking logic. If it is not 0, brake and decelerate according to the driver's operation.

[0042] According to the preset braking logic, after the current vehicle speed is braked to a specific speed, the vehicle reminds the driver that energy recovery has ended. Within 2 seconds, if the accelerator pedal opening and closing degree and the brake pedal opening and closing degree are still 0, energy recovery braking is performed according to the adjacent braking logic until the vehicle speed is braked to 0; if the accelerator pedal opening and closing degree and the brake pedal opening and closing degree are not 0, it is controlled by the driver.

[0043] Example 3

[0044] The electric vehicle energy recovery method based on machine learning in this embodiment is the same as that in Embodiment 1 or Embodiment 2 except that it is applied to reversing.

[0045] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for energy recovery of electric vehicles based on machine learning, comprising the following steps: Step 1, monitoring whether the opening and closing degree of the accelerator pedal of the electric vehicle is 0; Step 2: when the opening and closing degree of the accelerator pedal is monitored to be 0, obtaining vehicle driving information and driving environment information; Step 3, calculating the proximity items between the current vehicle driving information and the driving environment information and the preset braking logic; Step 4: Perform energy recovery braking using the adjacent braking logic; Step five, determine whether the current brake pedal opening degree is 0. If it is 0, brake until it stops according to the adjacent braking logic. If it is not 0, brake and decelerate according to the driver's operation. It is characterized in that the preset braking logic learns the driver's driving habits in different vehicle driving states and driving environments through a machine learning method, and obtains the braking logic for braking the vehicle from the current speed to a specific speed under different vehicle driving information and different driving environment information. After stopping the energy recovery braking, the vehicle prompts the driver that the energy recovery is over. Within 2 seconds after the vehicle prompts the driver that the energy recovery is over, if the accelerator pedal opening degree and the brake pedal opening degree are still 0, energy recovery braking is performed according to the adjacent braking logic until the vehicle speed is braked to 0.

2. The electric vehicle energy recovery method based on machine learning according to claim 1, characterized in that The vehicle driving information includes one or more of the vehicle gear position, current vehicle speed, and steering wheel rotation angle.

3. The electric vehicle energy recovery method based on machine learning according to claim 1 or 2, characterized in that: The driving environment information includes one or more of the following: whether there is a vehicle ahead and the distance, traffic light signals, the distance to the intersection ahead, whether there is a zebra crossing ahead, whether there is an obstacle ahead and the distance to the obstacle, and weather conditions.

4. The electric vehicle energy recovery method based on machine learning according to claim 3, characterized in that: In step three, a KNN algorithm, a regression algorithm, a random forest algorithm, or a SVM algorithm is used to calculate the current vehicle driving information, the driving environment information, and the adjacent items of the preset braking logic.

5. The electric vehicle energy recovery method based on machine learning according to claim 1, 2 or 4, characterized in that: The process of the driver manipulating the braking and deceleration in step 5 is used to correct the preset braking logic.

6. The electric vehicle energy recovery method based on machine learning according to claim 1, characterized in that: Within 2 seconds after the vehicle prompts the driver that energy recovery has ended, if the accelerator pedal opening and closing degrees and the brake pedal opening and closing degrees are not 0, then they are controlled by the driver.

7. Application of an electric vehicle energy recovery method based on machine learning in reversing according to any one of claims 1 to 6.

Images