Vehicle safety driving method and device

Abstract

The application discloses a vehicle safe driving method and device, and belongs to the technical field of intelligent driving. The method is applied to a vehicle controller and comprises the following steps: in the case that a to-be-tested vehicle is in an assisted driving state, in response to a manual takeover instruction, obtaining driving data of the to-be-tested vehicle; substituting the driving data of the to-be-tested vehicle into an intelligent decision model to obtain a decision result; and in response to the decision result being vehicle takeover, controlling the to-be-tested vehicle to keep the assisted driving state. The method intelligently decides the rationality of a manual takeover instruction in a driving process, avoids executing an incorrect manual takeover instruction, and ensures safe driving of the vehicle.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent driving, in particular to a vehicle safe driving method and device. BACKGROUND

[0002] At present, intelligent driving technology has been preliminarily popularized, has adaptive cruise and lane keeping capabilities, and has realized lane departure warning and lane departure correction functions, effectively reducing traffic safety accidents caused by driver negligence.

[0003] In the related art, when the vehicle is in an assisted driving state, if the driver touches the steering wheel or the accelerator pedal, the vehicle exits the assisted driving state and enters the manual takeover state. However, when the driver is unfamiliar with the road conditions, not concentrating or driving while tired, it is possible that the driver will mistakenly touch the steering wheel or the accelerator pedal, and the intelligent driving system will regard the human error as a reasonable manual takeover instruction, resulting in an accident that could have been avoided. SUMMARY

[0004] In view of this, the present application provides a vehicle safe driving method and device, which intelligently decides the rationality of the manual takeover instruction during driving, avoids executing incorrect manual takeover instructions, and ensures safe driving of the vehicle.

[0005] Specifically, the technical solutions include the following:

[0006] In one aspect, the present application provides a vehicle safe driving method applied to a vehicle control unit, the method comprising:

[0007] In the case that a to-be-tested vehicle is in an assisted driving state, in response to a manual takeover instruction, obtaining driving data of the to-be-tested vehicle;

[0008] Substituting the driving data of the to-be-tested vehicle into an intelligent decision-making model to obtain a decision result;

[0009] In response to the decision result being vehicle takeover, controlling the to-be-tested vehicle to remain in the assisted driving state.

[0010] In some embodiments, the driving data of the to-be-tested vehicle includes environmental data around the to-be-tested vehicle and vehicle data of the to-be-tested vehicle, and the obtaining of the driving data of the to-be-tested vehicle comprises:

[0011] Obtaining environmental data around the to-be-tested vehicle, wherein the environmental data around the to-be-tested vehicle includes an environmental model around the to-be-tested vehicle;

[0012] Obtaining vehicle data of the to-be-tested vehicle, wherein the vehicle data includes vehicle speed, steering wheel steering angle, and yaw angle speed.

[0013] In some embodiments, the acquiring the environmental data around the vehicle to be tested comprises:

[0014] acquiring an environmental image around the vehicle to be tested, the environmental image around the vehicle to be tested comprising a front image, a rear image, a left front side image, a right front side image, a left rear side image and a right rear side image of the vehicle to be tested;

[0015] modeling the environmental image around the vehicle to be tested to obtain an environmental model around the vehicle to be tested.

[0016] In some embodiments, the acquiring the driving data of the vehicle to be tested comprises:

[0017] acquiring a driving trajectory of the vehicle to be tested, wherein the driving trajectory is calculated according to the whole vehicle data of the vehicle to be tested within a preset time length.

[0018] In some embodiments, in the case that the vehicle to be tested is in an assisted driving state, before the acquiring the driving data of the vehicle to be tested in response to the manual takeover instruction, the method further comprises:

[0019] acquiring and storing the intelligent decision-making model, wherein the input and output of the intelligent decision-making model are driving data and decision-making results respectively, and the decision-making results comprise manual takeover and vehicle takeover.

[0020] In some embodiments, the acquiring the intelligent decision-making model comprises:

[0021] acquiring a set of driving data and a set of decision-making results, wherein the driving data and the decision-making results have a corresponding relationship;

[0022] training a target neural network model based on the set of driving data and the set of decision-making results to obtain the intelligent decision-making model.

[0023] In some embodiments, after the substituting the driving data of the vehicle to be tested into the intelligent decision-making model to obtain the decision-making results, the method further comprises:

[0024] in response to the decision-making results being manual takeover, controlling a warning component to perform a warning.

[0025] In some embodiments, the warning component comprises a multimedia device, a light device and a steering wheel, and the controlling the warning component to perform a warning comprises at least one of the following:

[0026] controlling the multimedia device to perform voice broadcast warning;

[0027] controlling the light device to flash;

[0028] controlling the steering wheel to vibrate or slightly horizontally twist.

[0029] In some embodiments, after the driving data of the vehicle under test is substituted into the intelligent decision model to obtain a decision result, the method further comprises:

[0030] In response to the decision result, a voice broadcast component is controlled to perform voice prompting.

[0031] In another aspect, the embodiments of the present application also provide a vehicle safe driving device, the device comprising:

[0032] An acquisition module is configured to, in a case where a vehicle under test is in an assisted driving state, acquire driving data of the vehicle under test in response to a manual takeover instruction;

[0033] A obtaining module is configured to substitute the driving data of the vehicle under test into an intelligent decision model to obtain a decision result;

[0034] A control module is configured to, in a case where the decision result is vehicle takeover, control the vehicle under test to remain in the assisted driving state.

[0035] The technical solutions provided by the embodiments of the present application have at least the following beneficial effects:

[0036] The vehicle safe driving method provided by the embodiments of the present application, when the vehicle under test is in the assisted driving state, acquires the driving data of the vehicle under test in response to the manual takeover instruction received by the vehicle controller, substitutes the driving data into the intelligent decision model to obtain a decision result, and in a case where the decision result is vehicle takeover, it is considered that the manual takeover instruction is a false operation at this time, so the manual takeover instruction can be ignored at this time, and the vehicle under test is controlled to remain in the assisted driving state, so as to automatically keep the vehicle in the original lane, so as to avoid the vehicle from executing the false manual takeover instruction, and ensure the safe driving of the vehicle. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0038] Figure 1 A method flowchart of a vehicle safe driving method provided by the embodiments of the present application;

[0039] Figure 2 A method flowchart of another vehicle safe driving method provided by the embodiments of the present application;

[0040] Figure 3A method flowchart for acquiring and storing an intelligent decision model in a vehicle safe driving method provided by an embodiment of the present application is provided.

[0041] Figure 4 A method flowchart for acquiring driving data of a vehicle to be tested in a vehicle safe driving method provided by an embodiment of the present application is provided.

[0042] Figure 5 A structural schematic diagram of a vehicle safe driving device provided by an embodiment of the present application is provided.

[0043] The above-described figures have shown the specific embodiments of the present application, which will be described in more detail hereinafter. These figures and textual descriptions are not intended to limit the scope of the concept of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0044] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort fall within the scope of protection of the present application.

[0045] Unless otherwise defined, all the technical terms used in the embodiments of the present application have the same meanings as commonly understood by those skilled in the art.

[0046] In order to make the technical solutions and advantages of the present application clearer, the embodiments of the present application will be described in further detail below with reference to the drawings.

[0047] At present, intelligent driving technology has been preliminarily popularized, has adaptive cruise and lane keeping capabilities, and has realized lane departure warning and lane departure correction functions, effectively reducing traffic safety accidents caused by driver negligence.

[0048] In the related art, when a vehicle is in an assisted driving state, if a driver touches a steering wheel or an accelerator pedal, the vehicle exits the assisted driving state and enters a manual takeover state. However, when the driver is unfamiliar with the road conditions, not focused, or fatigued, the driver may mistakenly touch the steering wheel or the accelerator pedal, and the intelligent driving system may regard the human error as a reasonable manual takeover instruction, resulting in an accident that could have been avoided.

[0049] In order to solve the technical problems in the related art, the embodiments of the present application provide a vehicle safe driving method and device, which intelligently decide the rationality of a manual takeover instruction during driving, avoid executing an incorrect manual takeover instruction, and ensure safe driving of the vehicle.

[0050] Figure 1 A method flowchart of a vehicle safe driving method provided by an embodiment of the present application is shown in FIG. 1. The method is applied to a vehicle controller and includes the following steps: Figure 1

[0051] In step 101, in response to a manual takeover instruction, driving data of the vehicle under test is obtained when the vehicle under test is in an assisted driving state.

[0052] In step 102, the driving data of the vehicle under test is substituted into an intelligent decision model to obtain a decision result.

[0053] In step 103, in response to the decision result being vehicle takeover, the vehicle under test is controlled to remain in the assisted driving state.

[0054] Therefore, the vehicle safe driving method provided by the embodiment of the present application, when the vehicle under test is in the assisted driving state, in response to the vehicle controller receiving the manual takeover instruction, the driving data of the vehicle under test is obtained, and the driving data is substituted into the intelligent decision model to obtain the decision result. In the case of the decision result being vehicle takeover, it is indicated that the manual takeover instruction is a false operation at this time, and thus the manual takeover instruction can be ignored at this time, the vehicle under test is controlled to remain in the assisted driving state, and the vehicle is automatically kept in the original lane, so as to avoid the vehicle executing the false manual takeover instruction and ensure safe driving of the vehicle.

[0055] In some embodiments, the driving data of the vehicle under test includes environmental data around the vehicle under test and vehicle data of the vehicle under test, and obtaining the driving data of the vehicle under test includes:

[0056] Obtaining the environmental data around the vehicle under test, wherein the environmental data around the vehicle under test includes an environmental model around the vehicle under test;

[0057] Obtaining the vehicle data of the vehicle under test, wherein the vehicle data includes vehicle speed, steering wheel steering angle, and yaw angle speed.

[0058] In some embodiments, obtaining the environmental data around the vehicle under test includes:

[0059] Obtaining environmental images around the vehicle under test, the environmental images around the vehicle under test including front images, rear images, left front side images, right front side images, left rear side images, and right rear side images of the vehicle under test;

[0060] Modeling the environmental images around the vehicle under test to obtain the environmental model around the vehicle under test.

[0061] In some embodiments, obtaining the driving data of the vehicle under test includes:

[0062] ​Obtaining a driving track of the vehicle to be tested, wherein the driving track is calculated according to whole vehicle data of the vehicle to be tested within a preset time length.

[0063] In some embodiments, in the case that the vehicle to be tested is in an auxiliary driving state, in response to the manual takeover instruction, before obtaining the driving data of the vehicle to be tested, the method further comprises:

[0064] Obtaining and storing an intelligent decision-making model, wherein the input and output of the intelligent decision-making model are driving data and decision-making results respectively, and the decision-making results include manual takeover and vehicle takeover.

[0065] In some embodiments, obtaining the intelligent decision-making model comprises:

[0066] Obtaining a driving data set and a decision-making result set, wherein the driving data and the decision-making results have a corresponding relationship;

[0067] Training a target neural network model based on the driving data set and the decision-making result set to obtain the intelligent decision-making model.

[0068] In some embodiments, after substituting the driving data of the vehicle to be tested into the intelligent decision-making model to obtain the decision-making results, the method further comprises:

[0069] In response to the decision-making results being manual takeover, controlling a warning component to perform a warning.

[0070] In some embodiments, the warning component includes a multimedia device, a light device, and a steering wheel, and controlling the warning component to perform a warning comprises at least one of the following:

[0071] Controlling the multimedia device to perform voice broadcast warning;

[0072] Controlling the light device to flash;

[0073] Controlling the steering wheel to vibrate or slightly horizontally twist.

[0074] In some embodiments, after substituting the driving data of the vehicle to be tested into the intelligent decision-making model to obtain the decision-making results, the method further comprises:

[0075] In response to the decision-making results, controlling a voice broadcast component to perform voice prompting.

[0076] Figure 2 Another method flowchart of a vehicle safety driving method provided by the embodiments of the present application, see Figure 2 The method is applied to a whole vehicle controller and comprises the following steps:

[0077] Step 201, obtaining and storing an intelligent decision-making model, wherein the input and output of the intelligent decision-making model are driving data and decision-making results respectively, and the decision-making results include manual takeover and vehicle takeover.

[0078] In some embodiments, the decision result further comprises a manual takeover correctness, wherein, in a case that the manual takeover correctness is less than or equal to a preset correctness, the decision result is vehicle takeover, and in this case, the vehicle control unit controls the vehicle to keep the assisted driving state; in a case that the manual takeover correctness is greater than the preset correctness, the decision result is divided into manual takeover or vehicle takeover.

[0079] That is, the manual takeover correctness can be a prerequisite for determining manual takeover or vehicle takeover.

[0080] In some embodiments, for example, the preset correctness can be 95%.

[0081] In some embodiments, referring to Figure 3 The step 201 comprises the following sub-steps:

[0082] Step 2011, obtaining a set of driving data and a set of decision results, wherein the driving data and the decision results have a corresponding relationship.

[0083] The set of driving data and the set of decision results having a corresponding relationship are obtained for training of the neural network model.

[0084] Step 2012, training a target neural network model based on the set of driving data and the set of decision results to obtain an intelligent decision model.

[0085] The target neural network model is trained by taking the set of driving data as model input and the set of decision results as model output, so as to obtain the intelligent decision model.

[0086] In some embodiments, the target neural network model comprises an image input layer, an image feature extraction layer, a data feature extraction layer, a feature fusion layer and an output layer, and the target neural network model has 1.2 million parameters.

[0087] In the embodiments of the present application, the set of driving data comprises environmental data around the vehicle, vehicle data and driving trajectory of the vehicle, wherein the environmental data around the vehicle comprises front image, rear image, left front side image, right front side image, left rear side image and right rear side image, and the vehicle data comprises vehicle speed, steering wheel steering angle and yaw rate.

[0088] The image input layer is used to receive the front image, the rear image, the left front side image, the right front side image, the left rear side image and the right rear side image, and the images of the six angles are usually 256x256 image data. Each image is processed by an independent preprocessing module for image data enhancement, including random rotation, scaling, shearing and color adjustment, to simulate various environmental images in the vehicle driving process.

[0089] The image feature extraction layer includes a plurality of convolutional networks for image feature extraction. During model training, image data of each angle is respectively extracted by a convolutional network configured with 32 and 64 filters. Each convolutional network includes a normalization and max-pooling layer to enhance the sensitivity of the target neural network model to image features.

[0090] The data feature extraction layer is composed of two fully connected networks for processing vehicle driving trajectory, vehicle speed, steering wheel steering angle, yaw angle speed and other features. Each fully connected network includes a normalization processing module to improve the processing efficiency and accuracy of the vehicle data.

[0091] The feature fusion layer is used to merge the image feature vector and the vehicle data feature vector by the Concatenate method. Specifically, the image feature vector and the vehicle data feature vector are merged by multiple dense connection layers (including a 512-unit ReLU layer and a Dropout layer) to form a comprehensive feature.

[0092] The output layer is used to output the decision result. Specifically, the artificial takeover accuracy is output by using a unit with a Sigmoid activation function, and "artificial takeover" or "vehicle takeover" is output by a Softmax layer.

[0093] In step 202, in the case that the vehicle to be tested is in an assisted driving state, the driving data of the vehicle to be tested is acquired in response to an artificial takeover instruction.

[0094] It can be understood that when the vehicle to be tested is in an assisted driving state, the driver does not need to operate the vehicle, and the vehicle can normally drive. If the vehicle controller receives an artificial takeover instruction from the driver, such as a steering instruction from the steering wheel or a brake or acceleration instruction from the pedal, it indicates that the driver may have the intention to artificially take over the vehicle. The driving data of the vehicle is acquired for subsequent calculation of the decision result.

[0095] In some embodiments, the driving data of the vehicle to be tested includes environmental data around the vehicle to be tested and vehicle data of the vehicle to be tested.

[0096] The environmental data around the vehicle can reflect the distribution of various traffic obstacles around the vehicle, and the vehicle data of the vehicle to be tested can reflect the speed, acceleration and driving direction of the vehicle. These data reflect the collision risk of the vehicle, and are used as inputs of the intelligent decision model to obtain the decision result.

[0097] In some embodiments, referring to Figure 4 , step 202 includes the following sub-steps:

[0098] In step 2021, the environment data around the vehicle to be tested is obtained, wherein the environment data around the vehicle to be tested comprises an environment model around the vehicle to be tested.

[0099] In some embodiments, step 2021 comprises: obtaining environment images around the vehicle to be tested, wherein the environment images around the vehicle to be tested comprise front images, rear images, left front side images, right front side images, left rear side images and right rear side images of the vehicle to be tested; and modeling the environment images around the vehicle to be tested to obtain the environment model around the vehicle to be tested.

[0100] By modeling the environment images around the vehicle to be tested, the environment model around the vehicle to be tested is obtained, which mainly comprises various types of recognized traffic information around the vehicle to be tested, wherein the traffic information comprises lane lines, vehicles, pedestrians, traffic lights, speed limit signs and traffic signboards, etc.

[0101] In step 2022, the vehicle data of the vehicle to be tested is obtained, wherein the vehicle data comprises vehicle speed, steering wheel steering angle and yaw angular velocity.

[0102] The vehicle data of the vehicle to be tested can reflect the vehicle speed, steering wheel steering angle and yaw angular velocity of the vehicle during driving, so the vehicle data of the vehicle is taken as the input of the intelligent decision-making model.

[0103] In step 2023, the driving trajectory of the vehicle to be tested is obtained, wherein the driving trajectory is calculated according to the vehicle data of the vehicle to be tested within a preset time period.

[0104] The driving trajectory of the vehicle to be tested is calculated according to the vehicle speed, steering wheel steering angle and yaw angular velocity of the vehicle to be tested, and is taken as the input of the intelligent decision-making model.

[0105] In step 203, the driving data of the vehicle to be tested is substituted into the intelligent decision-making model to obtain a decision result.

[0106] The environment model around the vehicle to be tested established according to the environment image data around the vehicle, the vehicle data of the vehicle to be tested and the driving trajectory of the vehicle to be tested are taken as the input of the intelligent decision-making model, and are substituted into the intelligent decision-making model, so as to obtain the decision result.

[0107] In some embodiments, when the decision result is the manual takeover accuracy, in the case that the manual takeover accuracy is less than or equal to a preset accuracy, for example, in the case that the manual takeover accuracy is less than or equal to 95%, the vehicle remains in the assisted driving state, and in the case that the manual takeover accuracy is greater than 95%, the next control step is made according to the manual takeover or vehicle takeover in the decision result.

[0108] Step 204, in response to the decision result, controlling the voice broadcast component to give a voice prompt.

[0109] That is, after obtaining the decision result, the voice broadcast component is controlled to give a voice prompt to the driver to remind the driver that the vehicle has the possibility of exiting the assisted driving state.

[0110] Step 205, in response to the decision result being vehicle takeover, controlling the vehicle under test to maintain the assisted driving state.

[0111] If the decision result is vehicle takeover, it indicates that if the vehicle responds to the manual takeover instruction, there will be a greater risk of collision, so the vehicle under test should be controlled to maintain the assisted driving state to avoid a collision accident.

[0112] Step 206, in response to the decision result being manual takeover, controlling the warning component to give a warning.

[0113] If the decision result is manual takeover, it indicates that if the vehicle responds to the manual takeover instruction, there will be no risk of collision, so the vehicle can respond to the manual takeover instruction, and at the same time, the warning component is controlled to give a warning to the driver and provide convenience for the driver to manually take over the vehicle.

[0114] In some embodiments, the warning component includes a multimedia device, a light device, and a steering wheel, and controlling the warning component to give a warning includes at least one of the following possible implementations:

[0115] In one possible implementation, the multimedia device is controlled to give a voice broadcast warning.

[0116] That is, in the case of the decision result being manual takeover, the multimedia device is controlled to give a voice broadcast warning to prompt the driver that the vehicle has exited the assisted driving state.

[0117] In another possible implementation, the light device is controlled to flash.

[0118] In the case of the decision result being manual takeover, the light device is controlled to flash to prompt the driver to pay attention to the fact that the vehicle has exited the assisted driving state.

[0119] In yet another possible implementation, the steering wheel is controlled to vibrate or vibrate laterally with a small amplitude.

[0120] In the case of the decision result being manual takeover, the driver is prompted by controlling the steering wheel to vibrate to warn the driver that the vehicle has exited the assisted driving state. At the same time, a small lateral torque is applied to the steering wheel to help the driver keep the vehicle straight, so as to facilitate the driver to manually take over the control of the vehicle after the vehicle exits the assisted driving state.

[0121] It can be understood that the above possible implementation manners can be performed alone, or in combination of two or three.

[0122] Therefore, the vehicle safe driving method provided by the embodiment of the present application, when the to-be-tested vehicle is in the assisted driving state, in response to the vehicle control unit receiving the manual takeover instruction, the driving data of the to-be-tested vehicle is acquired, and the driving data is substituted into the intelligent decision model to obtain a decision result. In the case of vehicle takeover, it is proved that the manual takeover instruction is a false operation at this time, so the manual takeover instruction can be ignored at this time, the to-be-tested vehicle is controlled to remain in the assisted driving state, and the vehicle is automatically kept in the original lane, so as to avoid the vehicle executing the false manual takeover instruction and ensure the safe driving of the vehicle.

[0123] Figure 5 A structural schematic diagram of a vehicle safe driving device provided by the embodiment of the present application is shown in Figure 5 The device 500 comprises:

[0124] The acquisition module 501 is configured to acquire the driving data of the to-be-tested vehicle in response to the manual takeover instruction when the to-be-tested vehicle is in the assisted driving state.

[0125] The obtaining module 502 is configured to substitute the driving data of the to-be-tested vehicle into the intelligent decision model to obtain a decision result.

[0126] The control module 503 is configured to control the to-be-tested vehicle to remain in the assisted driving state in response to the decision result being vehicle takeover.

[0127] In some embodiments, the acquisition module comprises:

[0128] The first acquisition submodule is configured to acquire the environmental data around the to-be-tested vehicle, wherein the environmental data around the to-be-tested vehicle comprises an environmental model around the to-be-tested vehicle.

[0129] The second acquisition submodule is configured to acquire the vehicle data of the to-be-tested vehicle, wherein the vehicle data comprises a vehicle speed, a steering wheel steering angle and a yaw angle speed.

[0130] In some embodiments, the first acquisition submodule comprises:

[0131] The third acquisition submodule is configured to acquire the environmental images around the to-be-tested vehicle, wherein the environmental images around the to-be-tested vehicle comprise a front image, a rear image, a left front side image, a right front side image, a left rear side image and a right rear side image of the to-be-tested vehicle.

[0132] The first obtaining submodule is configured to model the environmental images around the to-be-tested vehicle to obtain the environmental model around the to-be-tested vehicle.

[0133] In some embodiments, the acquisition module further comprises:

[0134] a fourth obtaining sub-module, configured to obtain a driving track of the vehicle to be tested, wherein the driving track is calculated according to the whole-vehicle data of the vehicle to be tested within a preset time length.

[0135] In some embodiments, the apparatus further includes:

[0136] a model obtaining module, configured to obtain and store an intelligent decision-making model, wherein the input and the output of the intelligent decision-making model are driving data and decision-making results respectively, and the decision-making results include manual takeover and vehicle takeover.

[0137] In some embodiments, the model obtaining module includes:

[0138] a fifth obtaining sub-module, configured to obtain a driving data set and a decision-making result set, wherein the driving data and the decision-making results have a corresponding relationship;

[0139] a second obtaining sub-module, configured to train a target neural network model based on the driving data set and the decision-making result set to obtain the intelligent decision-making model.

[0140] In some embodiments, the apparatus further includes:

[0141] a warning module, configured to control a warning component to perform warning in response to the decision-making results being manual takeover.

[0142] In some embodiments, the warning component includes a multimedia device, a light device and a steering wheel, and the warning module includes at least one of the following:

[0143] a first warning sub-module, configured to control the multimedia device to perform voice broadcast warning;

[0144] a second warning sub-module, configured to control the light device to flash;

[0145] a third warning sub-module, configured to control the steering wheel to vibrate or slightly horizontally twist.

[0146] In some embodiments, the apparatus further includes:

[0147] a prompting module, configured to control a voice broadcast component to perform voice prompting in response to the decision-making results.

[0148] Therefore, the vehicle safe driving device provided by the embodiment of the present application can obtain the driving data of the vehicle to be tested when the vehicle to be tested is in the assisted driving state, and the driving data is substituted into the intelligent decision model to obtain a decision result, and in the case that the decision result is that the vehicle is taken over, it is indicated that the manual takeover instruction is a false operation at this time, so the manual takeover instruction can be ignored at this time, the vehicle to be tested is controlled to keep the assisted driving state, and the vehicle to be tested is automatically kept in the original lane, so that the vehicle is prevented from executing the false manual takeover instruction, and safe driving of the vehicle is ensured.

[0149] In the present application, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. The term "a plurality of" refers to two or more, unless otherwise explicitly limited.

[0150] Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. The present application is intended to cover any variations, uses or adaptive changes of this application following the general principles thereof and including those expressly stated or implied herein. The specification and examples are considered exemplary only.

[0151] It should be understood that the present application is not limited to the precise construction that has been described above and shown in the accompanying drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present application is limited only by the claims appended hereto.

Claims

1. A method for safe driving of a vehicle, characterized in that, Applied to a vehicle controller, the method includes: Acquire and store an intelligent decision-making model, wherein the input and output of the intelligent decision-making model are driving data and decision results, respectively, and the decision results include manual takeover and vehicle takeover; When the vehicle under test is in assisted driving mode, in response to a manual takeover command, the driving data of the vehicle under test is acquired. The driving data of the vehicle under test is substituted into the intelligent decision-making model to obtain the decision result; In response to the decision result of vehicle takeover, the vehicle under test is controlled to maintain the assisted driving state; The driving data of the vehicle under test includes environmental data surrounding the vehicle and vehicle data. Obtaining the driving data of the vehicle under test includes: Acquire environmental data around the vehicle under test, wherein the environmental data around the vehicle under test includes an environmental model around the vehicle under test. Acquire the vehicle data of the vehicle under test, wherein the vehicle data includes vehicle speed, steering wheel angle and yaw rate; The intelligent decision-making model includes: Obtain a set of driving data and a set of decision results, wherein the driving data and the decision results have a corresponding relationship; The intelligent decision-making model is obtained by training a target neural network model based on the driving data set and the decision result set.

2. The vehicle safe driving method according to claim 1, characterized in that, The acquisition of environmental data around the vehicle under test includes: Acquire environmental images around the vehicle under test, including front, rear, left front, right front, left rear, and right rear images of the vehicle under test. The environmental image around the vehicle under test is modeled to obtain the environmental model around the vehicle under test.

3. The vehicle safe driving method according to claim 1, characterized in that, The acquisition of the driving data of the vehicle under test includes: The driving trajectory of the vehicle under test is obtained, wherein the driving trajectory is calculated based on the vehicle data of the vehicle under test within a preset time period.

4. The vehicle safe driving method according to claim 1, characterized in that, After substituting the driving data of the vehicle under test into the intelligent decision-making model to obtain the decision result, the method further includes: In response to the decision result, manual takeover is initiated, and the early warning component is controlled to issue an early warning.

5. The vehicle safe driving method according to claim 4, characterized in that, The warning component includes a multimedia device, a lighting device, and a steering wheel, and the control of the warning component to issue a warning includes at least one of the following: Control the multimedia device to broadcast warnings via voice; Control the flashing of the light device; Control the steering wheel vibration or slight lateral twisting.

6. The vehicle safe driving method according to claim 1, characterized in that, After substituting the driving data of the vehicle under test into the intelligent decision-making model to obtain the decision result, the method further includes: In response to the decision result, the voice broadcast component is controlled to provide voice prompts.

7. A vehicle safety driving device, characterized in that, The device includes: The model acquisition module is used to acquire and store intelligent decision-making models, wherein the input and output of the intelligent decision-making models are driving data and decision results, respectively, and the decision results include manual takeover and vehicle takeover. The acquisition module is used to acquire the driving data of the vehicle under test in response to a manual takeover command when the vehicle under test is in an assisted driving state. The module is used to input the driving data of the vehicle under test into the intelligent decision-making model to obtain the decision result; The control module is used to control the vehicle under test to maintain the assisted driving state in response to the decision result of vehicle takeover; The acquisition module includes: The first acquisition submodule is used to acquire environmental data around the vehicle under test, wherein the environmental data around the vehicle under test includes an environmental model around the vehicle under test. The second acquisition submodule is used to acquire the whole vehicle data of the vehicle under test, wherein the whole vehicle data includes vehicle speed, steering wheel angle and yaw rate; The model acquisition module includes: The fifth acquisition submodule is used to acquire a set of driving data and a set of decision results, wherein the driving data and the decision results have a corresponding relationship; The second submodule is used to train a target neural network model based on the driving data set and the decision result set to obtain the intelligent decision model.

Images