Vehicle auxiliary driving method, vehicle, medium and product

Abstract

The invention discloses a vehicle auxiliary driving method, a vehicle, a medium and a product, and relates to the technical field of vehicles, the method comprises the following steps: monitoring current driving state parameters of the vehicle in real time; the current driving state parameter is compared with an optimal driving strategy corresponding to a racing track where the vehicle is located to obtain a state difference, and the state difference is the difference between a local driving strategy parameter corresponding to the current position of the vehicle in the optimal driving strategy and the current driving state parameter; and outputting vehicle correction prompt information according to the state difference. Compared with a traditional training scheme, the training mode based on real-time data analysis and dynamic prompt can greatly shorten the growth cycle of the driver, and the training efficiency and quality are improved.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a vehicle assisted driving method, vehicle, medium and product. Background Art

[0002] Current optimization of track driving techniques relies heavily on the driver's subjective experience and manual guidance from a coach. Traditional approaches to assisting drivers in improving their driving skills also have significant limitations. For example, traditional approaches primarily rely on post-race or training sessions to review and optimize driving techniques using recorded racing data (such as track heat maps generated by collecting basic parameters like vehicle speed and engine speed). Consequently, traditional training programs can only provide a static reference and exhibit significant lag, resulting in poor training effectiveness.

[0003] The above content is only used to assist in understanding the technical solution of this application and does not constitute an admission that the above content is prior art. Summary of the Invention

[0004] The main purpose of this application is to provide a vehicle assisted driving method, vehicle, medium and product, aiming to solve the technical problem that traditional training programs can only provide static references, have obvious lags, lack real-time dynamic guidance value, and thus lead to poor training results.

[0005] To achieve the above objectives, the present application proposes a vehicle assisted driving method, which includes the following steps:

[0006] Real-time monitoring of current driving state parameters of the vehicle;

[0007] Comparing the current driving state parameter with the optimal driving strategy corresponding to the track where the vehicle is located to obtain a state difference, wherein the state difference is the difference between the local driving strategy parameter corresponding to the current position of the vehicle in the optimal driving strategy and the current driving state parameter;

[0008] According to the state difference, vehicle correction prompt information is output.

[0009] Optionally, before the step of monitoring the current driving state parameters of the vehicle in real time, the method further comprises:

[0010] Constructing a digital track model simulating the track and the dynamic constraints of the vehicle;

[0011] The shortest time for the vehicle to complete the track is set as a simulation goal. Under the constraints of the dynamic constraints, the driving process of the vehicle in the digital track model is simulated to generate an optimal driving strategy corresponding to the complete track.

[0012] Optionally, the step of constructing a digital track model simulating the track and the dynamic constraints of the vehicle includes:

[0013] Acquiring basic geometric data of the track, environmental data of the track, and a priori driving parameters, wherein the a priori driving parameters are collected by a priori vehicle traveling on the track;

[0014] Building an initial model of the track based on the positioning data and posture data in the prior driving parameters;

[0015] Extracting physical property features of the track from the geometric basic data and the environmental data, and loading the physical property features into the initial model to obtain the digital track model;

[0016] The maximum acceleration and braking performance of the vehicle in the digital track model are calculated according to the dynamic parameters of the vehicle to obtain the dynamic constraint conditions.

[0017] Optionally, the step of taking the shortest time for the vehicle to complete the track as a simulation goal and simulating the vehicle's driving process in the digital track model under the constraints of the dynamic constraints to generate an optimal driving strategy corresponding to the complete track includes:

[0018] Constructing a reward function with the goal of minimizing the time it takes for the vehicle to complete the track;

[0019] Under the constraints of the dynamic constraints, simulating the driving process of the vehicle in the digital track model based on the current driving strategy to obtain an experience sample set, and adding the experience sample set to an experience replay pool, wherein the experience samples in the experience sample set consist of a node vehicle state, a node vehicle action corresponding to the node vehicle state in the current driving strategy, a reward value of the node vehicle action determined by the reward function, and a next node vehicle state of the node vehicle state;

[0020] Collecting experience samples from the experience replay pool, updating network parameters of the strategy generation network, and generating a new current driving strategy through the updated strategy generation network;

[0021] Based on the new current driving strategy, the step of simulating the driving process of the vehicle in the digital track model based on the current driving strategy under the constraints of the dynamic constraints to obtain an experience sample set is returned to be executed until the strategy generation network converges, and the driving strategy generated by the converged strategy generation network is used as the complete optimal driving strategy corresponding to the track.

[0022] Optionally, before the step of constructing a digital track model simulating the track and the dynamic constraints of the vehicle, the method further includes:

[0023] If a driving strategy of the same model as that of the vehicle exists in a preset strategy library corresponding to the track, the optimal driving strategy is obtained by matching the driving strategies of the same model as that of the vehicle based on the vehicle state;

[0024] If there is no driving strategy for the same model as that of the vehicle in the preset strategy library corresponding to the track, calculating the similarity between the vehicle and the models corresponding to each driving strategy in the preset strategy library based on the vehicle parameters of the vehicle;

[0025] When the maximum similarity among the similarities is less than or equal to a preset threshold, executing the step of constructing a digital track model simulating the track and the dynamic constraint conditions of the vehicle;

[0026] When a maximum similarity among the similarities is greater than a preset threshold, an optimal driving strategy for the vehicle is determined based on the driving strategy of the vehicle model corresponding to the maximum similarity.

[0027] Optionally, the vehicle assisted driving method further includes:

[0028] Monitoring the vehicle state of the vehicle, wherein the vehicle state includes the vehicle's own attribute state and the vehicle's environment state;

[0029] After the change in the vehicle state of the vehicle exceeds a change threshold, the step of matching the vehicle state of the vehicle with the driving strategies of the same model as the vehicle to obtain the optimal driving strategy is returned based on the changed vehicle state.

[0030] Optionally, the virtual path in the optimal driving strategy and the driving parameters on the virtual path are projected into the field of view of the vehicle driver, the vehicle correction prompt information includes at least one of an oil brake operation prompt and a steering operation prompt, and the step of outputting the vehicle correction prompt information according to the state difference includes:

[0031] If the state difference includes a speed difference, generating the oil brake operation prompt based on the speed difference and superimposing the oil brake operation prompt on the virtual path, wherein the oil brake operation prompt includes an accelerator action or a brake action, and an opening corresponding to the action;

[0032] In a case where the state difference includes a position difference, generating the steering operation prompt based on the position difference, wherein the steering operation prompt includes a steering direction and a steering angle;

[0033] By controlling the vibration of different positions on the steering wheel of the vehicle, the steering direction is prompted, and by controlling the vibration of the steering wheel to stop, it is prompted that the steering angle has been reached.

[0034] In addition, to achieve the above-mentioned purpose, the present application also proposes a vehicle, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the vehicle assisted driving method as described above.

[0035] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium. A computer program is stored on the storage medium, and when the computer program is executed by the processor, the steps of the vehicle assisted driving method described above are implemented.

[0036] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the vehicle assisted driving method described above.

[0037] One or more technical solutions proposed in this application have at least the following technical effects:

[0038] In an embodiment of the present application, the current driving state parameters of the vehicle are monitored in real time; the current driving state parameters are compared with the optimal driving strategy corresponding to the track where the vehicle is located to obtain a state difference, wherein the state difference is the difference between the local driving strategy parameters corresponding to the current position of the vehicle in the optimal driving strategy and the current driving state parameters; according to the state difference, vehicle correction prompt information is output. It can be understood that the embodiment of the present application monitors the current driving state parameters of the vehicle in real time and compares them with the difference between the local driving strategy parameters corresponding to the current position of the vehicle in the optimal driving strategy, thereby generating prompts based on the differences and outputting vehicle correction prompt information in real time, which can guide the driver to adjust the driving behavior in time and make the vehicle gradually approach the optimal driving state. This real-time dynamic guidance can significantly improve the training effect compared to the post-review in the traditional training program, help the driver master the correct driving skills faster, and improve the performance on the track. Therefore, the training method based on real-time data analysis and dynamic prompts in the embodiment of the present application can greatly shorten the driver's growth cycle and improve the training efficiency and quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0040] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0041] Figure 1 This is a flow chart of the first embodiment of the vehicle assisted driving method of the present application;

[0042] Figure 2 This is a flow chart of the second embodiment of the vehicle assisted driving method of the present application;

[0043] Figure 3 This is a schematic diagram of the data processing framework in the vehicle assisted driving method of this application;

[0044] Figure 4 A schematic diagram of the process of strategy migration in the vehicle assisted driving method of this application;

[0045] Figure 5 This is a schematic diagram of a scenario in the vehicle assisted driving method of this application;

[0046] Figure 6 This is a schematic diagram of the device structure of the hardware operating environment involved in the vehicle assisted driving method in the embodiment of the present application.

[0047] The purpose, features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0048] It should be understood that the specific embodiments described herein are merely used to explain the technical solutions of the present application and are not intended to limit the present application.

[0049] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0050] Current optimization of track driving techniques relies heavily on the driver's subjective experience and manual guidance from a coach. Traditional approaches to assisting drivers in improving their driving skills also have significant limitations. For example, traditional approaches primarily rely on post-race or training sessions to review and optimize driving techniques using recorded racing data (e.g., track heat maps generated by collecting basic parameters such as vehicle speed and engine speed). Consequently, traditional training programs can only provide a static reference, exhibit significant lags, and lack real-time, dynamic guidance, resulting in poor training effectiveness.

[0051] The main solution of the embodiment of the present application is: real-time monitoring of the current driving state parameters of the vehicle; comparing the current driving state parameters with the optimal driving strategy corresponding to the track where the vehicle is located to obtain a state difference, wherein the state difference is the difference between the local driving strategy parameters corresponding to the current position of the vehicle in the optimal driving strategy and the current driving state parameters; and outputting vehicle correction prompt information based on the state difference.

[0052] It is understandable that the embodiment of the present application monitors the current driving state parameters of the vehicle in real time and compares the difference between the parameters and the local driving strategy parameters corresponding to the current position of the vehicle in the optimal driving strategy, thereby generating prompts based on the differences and outputting vehicle correction prompt information in real time, which can guide the driver to adjust the driving behavior in time and make the vehicle gradually approach the optimal driving state. This real-time dynamic guidance can significantly improve the training effect, help the driver master the correct driving skills faster, and improve the performance on the track. Compared with traditional training programs, the training method based on real-time data analysis and dynamic prompts in the embodiment of the present application can greatly shorten the driver's growth cycle and improve training efficiency and quality.

[0053] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a cloud platform, computer, mobile phone, etc., or a vehicle that can realize the above functions.

[0054] Based on the above introduction, the embodiment of the present application provides a vehicle assisted driving method, referring to Figure 1 , which is a flow chart of the first embodiment of the vehicle assisted driving method of the present application.

[0055] In this embodiment, the vehicle assisted driving method includes steps S10 to S30:

[0056] Step S10, real-time monitoring of the vehicle's current driving state parameters;

[0057] It should be noted that, in this embodiment, the implementing body of the above-mentioned vehicle assisted driving method can be a vehicle, a cloud server that communicates with the vehicle in real time, or a car computer configured on the vehicle, etc. For the convenience of description, the car computer will be used as an example in the subsequent description. After the vehicle positioning data detects that the vehicle has entered the track, the vehicle assisted driving function can be automatically triggered to turn on, or the assisted driving function can be triggered to turn on under the operation of the driver. It is worth noting that after the vehicle assisted driving function is turned on, the vehicle can first match the optimal driving strategy of the track it is on to prepare for vehicle assisted driving. Among them, the optimal driving strategy is a strategy pre-configured according to the track, which can include the optimal driving route, and the driving speed at each position on the driving route, etc.

[0058] For example, after the vehicle's assisted driving function is enabled and the driver begins driving on a track, the vehicle's onboard computer will monitor the vehicle's current driving state in real time. The current driving state typically includes the vehicle's position on the track, which can be the longitudinal position of the lane, for example, the distance relative to the lane markings. Additionally, the current vehicle state may include the vehicle's speed and acceleration in various directions. For example, monitoring the vehicle's current driving state can be accomplished through various sensor systems onboard the vehicle, which collect a variety of data from the vehicle in real time and in all directions. These sensors include, but are not limited to, speed sensors, accelerometers, gyroscopes, GPS positioning systems, and cameras. Speed ​​sensors accurately measure the vehicle's current speed; accelerometers monitor changes in the vehicle's acceleration in various directions, such as longitudinal acceleration (acceleration or deceleration in the forward or reverse direction) and lateral acceleration (lateral acceleration when turning); gyroscopes determine the vehicle's attitude information, such as pitch, roll, and yaw angles; GPS positioning systems provide precise coordinates of the vehicle's position on the track; and cameras capture information about the vehicle's surroundings, such as the track boundaries. The vehicle computer integrates, pre-processes and analyzes the raw data collected by these sensors to build a complete data model of the vehicle's current driving status parameters, which facilitates the subsequent assisted driving steps of the original driving vehicle.

[0059] Step S20: comparing the current driving state parameters with the optimal driving strategy corresponding to the track where the vehicle is located to obtain a state difference, wherein the state difference is the difference between the local driving strategy parameters corresponding to the current position of the vehicle in the optimal driving strategy and the current driving state parameters;

[0060] It should be noted that after the vehicle starts the assisted driving function, the vehicle's computer can automatically match the optimal driving strategy corresponding to the current track. In this embodiment, the optimal driving strategy is generated by machine learning, combined with a large amount of historical data, track characteristic analysis, vehicle performance parameters, professional prior driving experience and other factors, to be the most ideal driving plan for a specific track. It can exist in the form of a mathematical model or a data table, which describes in detail the optimal driving parameters that the vehicle should achieve at each position on the track, including the optimal speed curve, the optimal braking point, the optimal steering timing and angle, the optimal throttle control strategy, etc., and these parameters can finally form the optimal path. (sequence of spatial coordinate points) and recommended velocity field V *(s) (the optimal speed at each position) to constitute the driving strategy parameters, and correspondingly, the above-mentioned local driving strategy parameters refer to the driving strategy parameters corresponding to a certain position of the track in the optimal driving strategy. For example, for a certain specific curve, the optimal driving strategy will clearly stipulate the ideal speed of the vehicle at the entrance of the curve, the brake start position, the brake force variation curve, the variation law of the steering wheel rotation angle with time, etc., so as to ensure that the vehicle can pass through the curve in the shortest time and the safest way, while maximizing the use of the advantages of the track.

[0061] For example, the current driving state parameters are a set of comprehensive data collected by the vehicle at a specific moment based on real-time sensor acquisition and processing. These data comprehensively reflect the actual running conditions of the vehicle on the track, and can cover the dynamic characteristics of the vehicle (such as speed, acceleration, steering angle, etc.) and position information (such as GPS coordinates, relative position on the track, etc.). For example, when the vehicle drives to a certain curve, the current driving state parameters can include the entry speed of the vehicle, the brake force, the steering wheel rotation angle, the lateral offset of the vehicle in the curve, and other specific parameters.

[0062] In the comparison process: the vehicle can first locate the local driving strategy parameters corresponding to the current position in the data model of the optimal driving strategy according to the current position information of the vehicle. Then, the local driving strategy parameters are compared and analyzed one by one with the corresponding actual parameters in the current driving state parameters. For example, the difference between the actual speed of the vehicle at a certain point on the track and the optimal speed stipulated in the optimal driving strategy at that point is compared; the difference between the actual brake force and the optimal brake force is compared; the actual steering angle is compared with the optimal steering angle, etc. Through this detailed comparison and analysis, the state difference between the two is calculated, which can include speed difference, acceleration difference, steering angle difference, and trajectory deviation.

[0063] Step S30, outputting the vehicle correction prompt information according to the state difference.

[0064] For example, after the vehicle computer acquires state difference data, it can analyze and calculate these differences using preset algorithms and rules. Based on the type and degree of the difference, corresponding vehicle correction prompts are generated. (In actual applications, the correction sensitivity can be set as needed, and the difference threshold that triggers the correction sensitivity prompt can be set accordingly. That is, the prompt information is only generated when the state difference exceeds the difference threshold.) These prompts can be presented in various forms, such as voice prompts, text and graphic prompts on the vehicle display, AR (Augmented Reality) or head-up display, and steering wheel vibration feedback. The prompts are specific and targeted, clearly informing the driver which driving parameters or operations need to be adjusted, as well as the direction and approximate magnitude of the adjustment. For example, if the state difference indicates that the vehicle's current speed is too high, the prompt message may be "Current speed is too high. Please slow down appropriately. The expected deceleration is 10-15 km / h." If the vehicle's steering angle is insufficient at a curve, the prompt message may be "Insufficient steering angle at the curve. Please increase the steering wheel angle by approximately 5-10 degrees." In addition, the car computer can also reasonably arrange the timing of prompts according to the vehicle's current track position and driving stage to ensure that the prompt information will not interfere with the driver's normal driving operations and can take effect in a timely manner.

[0065] In this embodiment, the current driving state parameters of the vehicle are monitored in real time; the current driving state parameters are compared with the optimal driving strategy corresponding to the track where the vehicle is located to obtain a state difference, wherein the state difference is the difference between the local driving strategy parameters corresponding to the current position of the vehicle in the optimal driving strategy and the current driving state parameters; based on the state difference, vehicle correction prompt information is output. It can be understood that the embodiment of the present application monitors the current driving state parameters of the vehicle in real time and compares them with the local driving strategy parameters corresponding to the current position of the vehicle in the optimal driving strategy, thereby generating prompts based on the differences and outputting vehicle correction prompt information in real time, which can guide the driver to adjust his driving behavior in a timely manner so that the vehicle gradually approaches the optimal driving state. This real-time dynamic guidance can significantly improve the training effect, help the driver master the correct driving skills more quickly, and improve performance on the track. Compared with traditional training programs, the training method based on real-time data analysis and dynamic prompts in the embodiment of the present application can greatly shorten the driver's growth cycle and improve training efficiency and quality.

[0066] In a feasible embodiment, a virtual path in the optimal driving strategy and driving parameters on the virtual path are projected into the field of view of the vehicle driver. The vehicle correction prompt information includes at least one of an oil brake operation prompt and a steering operation prompt. The step of outputting the vehicle correction prompt information according to the state difference includes steps S31 to S33:

[0067] Step S31: if the state difference includes a speed difference, generating a brake operation prompt based on the speed difference and superimposing the brake operation prompt on the virtual path, wherein the brake operation prompt includes an accelerator action or a brake action, and an opening corresponding to the action;

[0068] It should be noted that the optimal driving strategy may include a virtual path and driving parameters on the virtual path, and the driving parameters may be the driving speed at each position on the virtual path, or may include driving operations at each position, such as braking, accelerator, and steering. The virtual path and the driving parameters on the virtual path can be projected into the field of view of the vehicle driver. For example, the optimal driving strategy can be projected through a head-up display configured on the vehicle, or AR glasses that are communicatively connected to the vehicle's onboard computer. In addition, for the vehicle correction prompt information, it may include at least one of an oil brake operation prompt and a steering operation prompt, such as only an oil brake operation prompt, or only a steering operation prompt, or both an oil brake operation prompt and a steering operation prompt, depending on the specific operation of the driver.

[0069] For example, if the state difference includes a speed difference, under normal circumstances, the speed difference may manifest as the current speed being higher or lower than the specified speed at the corresponding location in the optimal driving strategy. For example, on a straight road section, the optimal driving strategy requires the vehicle to maintain a speed of 150 km / h, but the current vehicle speed is 160 km / h. The speed difference is +10 km / h; if the actual speed is 140 km / h, the speed difference is -10 km / h. Based on the speed difference, a corresponding brake operation prompt is generated. If the speed difference is positive (the current speed is higher than the optimal speed), the prompt typically includes the braking action and the corresponding brake position. For example, based on a speed difference of +10 km / h, the machine calculates that the brake pedal needs to be depressed 30% to gradually reduce the vehicle speed to the optimal speed range. If the speed difference is negative (the current speed is lower than the optimal speed), the prompt includes the accelerator action and the corresponding accelerator position. For example, if the speed difference is -10 km / h, the driver is prompted to depress the accelerator pedal 20% to increase vehicle power output and speed. It should be noted that the specific rules for generating the oil brake operation prompt can be set by technicians according to actual needs and vehicle performance, and will not be repeated here.

[0070] In addition, the generated oil and brake operation prompt information is superimposed on the virtual path within the driver's field of view in the form of intuitive graphics, text or symbols. The virtual path is the ideal driving trajectory of the vehicle generated by the machine based on the optimal driving strategy simulation. After superimposing the prompt information, when the driver observes the virtual path, he can simultaneously see the throttle or brake operation and the opening degree that should be performed at different positions on the path. For example, on the virtual path, the braking action is indicated by a red arrow at the position corresponding to the braking position, and the brake opening value (such as 30%) is marked next to the arrow; at the position where acceleration is required, the throttle action is indicated by a green arrow, and the throttle opening value (such as 20%) is marked.

[0071] Step S32: if the state difference includes a position difference, generating a steering operation prompt based on the position difference, wherein the steering operation prompt includes a steering direction and a steering angle;

[0072] For example, by comparing the current vehicle position with the virtual path of the corresponding position in the optimal driving strategy, the position difference is accurately calculated. The position difference may be manifested as a lateral offset of the vehicle relative to the virtual path, that is, the vehicle's driving trajectory deviates from the optimal path. For example, at a certain curve, the actual driving trajectory of the vehicle deviates by 0.5 meters to the outside of the curve, then the 0.5-meter offset to the outside of the curve is the position difference. Based on the position difference, the vehicle computer can use a pre-set algorithm to calculate the steering operation prompt required to correct the position deviation. The steering operation prompt can indicate the steering direction (left or right) and the steering angle. For example, if the vehicle deviates to the right, the steering direction is prompted to be left; based on the size of the position difference and the driving dynamics characteristics of the vehicle, the steering angle of 15 degrees to the left is calculated to gradually return the vehicle to the virtual path.

[0073] Step S33, by controlling the vibration of different positions on the steering wheel of the vehicle, the steering direction is prompted, and by controlling the vibration of the steering wheel to stop, it is prompted that the steering angle has been reached.

[0074] It should be noted that, in order to ensure that the prompts are more intuitive, steering operation prompts can be given through the vibration of the steering wheel.

[0075] For example, based on the steering direction indicated in the generated steering operation prompt, vibration is controlled in the corresponding area of ​​the steering wheel. For example, if the steering direction indicated is left, vibration is controlled in the left area of ​​the steering wheel; if the steering direction is right, vibration is controlled in the right area of ​​the steering wheel. This vibration prompt method intuitively informs the driver which direction to turn the steering wheel by different vibration positions. After the driver begins turning the steering wheel, the system monitors the vehicle's steering angle changes in real time. When the vehicle reaches the required steering angle, the system immediately controls the steering wheel vibration to stop. For example, if the steering angle indicated is 15 degrees, when the vehicle actually reaches 15 degrees, the steering wheel vibration stops instantly, indicating to the driver that the target steering angle has been reached. It is also understood that steering operation prompts can also be superimposed on the virtual path in the form of images and text to provide prompts to the driver, which will not be further explained here.

[0076] In order to explain this case clearly, it will be explained with reference to actual scenarios.

[0077] The car will obtain the vehicle status (such as speed, location, direction, etc.) in real time and will also use the recommended best route and the optimal speed V at each position * (s). For example, by using AR glasses or in-vehicle displays, drivers can be provided with visual route and operation suggestions.

[0078] First, the vehicle system calculates the gap (deviation) between the current vehicle and the recommended path, speed, and operation, and then generates operation suggestions based on these deviations.

[0079] These suggestions will be dynamically displayed to the driver through AR glasses or HUD (Head Up Display), including recommended driving routes, speeds, operating parameters and warning information.

[0080] The vehicle system will also monitor key parameters such as tire pressure and temperature in real time, and will automatically adjust the strategy if any abnormality is detected.

[0081] For more biased generation operations, we recommend:

[0082] Deviation calculation: Path deviation, the distance difference between the current vehicle position and the recommended path. Speed ​​deviation, the difference between the current speed and the recommended speed. Steering wheel angle deviation, the difference between the current steering wheel angle and the recommended angle.

[0083] Action Suggestions: If the path deviates significantly, steering angle adjustments are recommended. The system dynamically adjusts throttle, brake, and steering based on the deviation. Speed ​​Control Suggestions adjust the throttle or brake based on the optimal speed. Steering Suggestions adjust based on the recommended steering wheel angle.

[0084] Reference Figure 2 For the first embodiment of the vehicle auxiliary driving method based on the present application, the flowchart of the second embodiment of the vehicle auxiliary driving method proposed in the present application is shown. The same or similar contents in this embodiment as the above embodiments can be referred to the above introduction, and will not be described in detail hereinafter. Before the step of monitoring the current driving state parameters of the vehicle in real time, the method further comprises steps S100-S200:

[0085] Step S100, constructing a digital track model simulating a track and a dynamic constraint condition of the vehicle;

[0086] Step S200, taking the shortest time for the vehicle to complete the track as the simulation target, simulating the driving process of the vehicle in the digital track model under the constraint of the dynamic constraint condition, to generate a complete optimal driving strategy corresponding to the track.

[0087] For example, a digital track model simulating a real track can be built by collecting detailed information of the track in advance. For example, a vehicle equipped with a high-precision laser radar can be used to scan the track in all directions to obtain three-dimensional point cloud data of the track surface. These point cloud data can contain accurate information such as the height, slope, and curvature of the track. In addition, high-resolution satellite images and unmanned aerial vehicle aerial images can also be combined to further supplement the surrounding environment information of the track, such as the location of the track boundary, buffer zone, and landmark objects (such as corner warning signs, maintenance area entrances, etc.). The specific data content collected by the technical personnel can be set according to the actual needs, and will not be described here. After the data collection is completed, the physical characteristics of the track represented by the data can be converted into a digital three-dimensional model. The model can restore the actual shape and size of the track. At the same time, in order to ensure the authenticity, the surface characteristics under different weather conditions such as dry, wet, and snowy can also be added to the track, so as to reflect the influence of the actual track weather on the vehicle driving.

[0088] The dynamic constraint condition of the vehicle is a mathematical quantitative description of the physical characteristics and motion rules of the vehicle. For example, it can be constructed from multiple key parameters of the vehicle. In terms of the power system, the power-speed characteristic curve and torque output characteristic of the engine can be collected to determine the acceleration performance of the vehicle under different gears. In terms of tire performance, the parameters such as tire grip force and cornering stiffness are studied, which change with the temperature, air pressure, wear degree of the tire and the road condition. For example, when the tire temperature rises, the grip force will change, and the machine will adjust the friction force constraint between the tire and the road accordingly to ensure the accuracy of the simulation. It can be understood that the dynamic constraint condition of the vehicle is mainly used to constrain the motion performance of the vehicle in the simulated environment.

[0089] After constructing the model and constraints, the process of a vehicle driving on the digital track model can be simulated. During the simulation, the core objective is to minimize the time it takes for the vehicle to complete the track. Accordingly, during the simulation, a reinforcement learning algorithm can continuously try the optimal driving strategy to ensure that the vehicle traverses the entire track at the fastest speed while satisfying all dynamic constraints. For example, at the beginning of the simulation, a series of initial parameters can be set for the vehicle, such as initial position, initial speed, and initial gear. Then, based on the vehicle's dynamic constraints, the vehicle's motion state changes within each time step are calculated. During the simulation, the agent corresponding to the reinforcement learning algorithm can continuously try different driving strategies. For example, when entering a curve, it simulates different combinations of braking timing, braking force, steering angle, and steering speed, observing the vehicle's trajectory, speed changes, and acceleration performance when exiting the curve. Through continuous iteration and optimization, the agent ultimately obtains the optimal driving strategy for the entire track.

[0090] It can be understood that in this embodiment, the optimal driving strategy under ideal conditions is found through iterative optimization through simulation, and is used as an auxiliary basis for training the driver's driving skills, thereby ensuring the rationality of the prompt information output in the vehicle assisted driving method of this application and the training effect of the driver.

[0091] In a feasible embodiment, the step of constructing a digital track model of a simulated track and the dynamic constraints of the vehicle includes steps S110 to S140:

[0092] Step S110, obtaining basic geometric data of the track, environmental data of the track, and a priori driving parameters, wherein the a priori driving parameters are data collected while the vehicle is driving on the track;

[0093] Step S120, building an initial model of the track based on the positioning data and posture data in the prior driving parameters;

[0094] Step S130, determining the physical property characteristics of the initial model using the geometric basic data and the environmental data, and loading the physical property characteristics into the initial model to obtain a digital track model;

[0095] In step S140 , the maximum acceleration and braking performance of the vehicle in the digital track model are calculated based on the vehicle's dynamic parameters to obtain dynamic constraint conditions.

[0096] Exemplarily, the geometric basic data of the track can include the width, height, curvature, slope, etc. of the track, which can be measured at key positions of the track. The environmental data of the track includes meteorological data such as temperature, humidity, wind force, road surface temperature, etc., and can also include features around the track. The prior driving parameters refer to the data collected by various sensors loaded on the vehicle during the prior driving of the vehicle on the track, for example, point cloud data collected by a laser radar, positioning data collected by a GPS (Global Positioning System), and attitude data collected by an IMU (Inertial Measurement Unit). In addition, the prior driving parameters can also include parameters of the vehicle itself, such as tire temperature, tire pressure, torque, throttle, brake, steering wheel angle, G value, etc.

[0097] The above-mentioned multi-source data collected are uniformly time-stamped and synchronized to form a collection In the formula, N refers to the length of the data collection, wherein d i is a data collection of different sources at time t i . All data are synchronized by uniform time stamping, and a sliding window buffer is used to ensure data integrity and time sequence consistency. Exemplarily, time synchronization can include PTP (Precision Time Protocol) or GPS PPS (Global Positioning System One Pulse Per Second) signal calibration, and all data packets are attached with high-precision time stamp t. Spatial alignment: through an external parameter matrix Unify each sensor data to the vehicle coordinate system:

[0098]

[0099] In the formula, is each sensor data before alignment, is an external parameter matrix, is each sensor data after alignment. The data after space-time synchronization is subjected to sliding window processing: the window length T w , the data in [t-T w , t] is buffered and interpolated, for example, the missing data is filled by interpolation method. The collected data is preprocessed, and the processing method can include removing outliers, smoothing filtering, normalization, etc.

[0100] After the data acquisition is completed, the digital track model is constructed based on the acquired multi-source data. For example, the initial model of the track can be constructed based on the positioning data and posture data in the prior driving parameters. For example, the spatial shape of the track will be constructed using the GPS three-dimensional point cloud in the positioning data, and N GPS points g will be collected. i =[x i ,y i ,z i ], then fills in the 3D point cloud data based on the driving trajectory in the positioning data to construct the spatial shape of the track. At the same time, attitude data collected by the IMU can be used for attitude compensation to correct errors caused by sensor drift. After attitude compensation, the initial model of the track can be obtained.

[0101] Then, the physical properties of the initial model are calculated using the geometric basic data and environmental data. For example, the calculated physical properties include:

[0102] Curvature: Where v and a0 are the first-order derivative and second-order derivative of the spatial curve corresponding to the curve, respectively, which are used to describe the curvature of the track and affect the force on the vehicle when turning.

[0103] slope: dz is the height change in the z-axis direction, ds is the unit distance of the track, and the slope is used to describe the uphill and downhill conditions of the track, which affects vehicle power and safety.

[0104] Friction coefficient: μ = f(T tire (tire temperature, humidity, road temperature), taking into account factors such as tires, humidity, and road temperature, and reflects the vehicle's grip.

[0105] It should be noted that the above curvature, slope and friction coefficient can be formed by combining the geometric basic data and environmental data with the above three-dimensional point cloud data. After fusion, Kalman filtering is used to remove the noise and abnormal points in the above data to make the curve of the physical attribute characteristics smoother. The obtained physical attribute characteristics are loaded into the initial model to obtain the digital track model. Figure 3 , which is a schematic diagram of the data processing framework in the embodiment of the present application. As shown in the figure, in this embodiment, there are various data used to build the digital track model, such as GPS data obtained by the positioning sensor, posture data obtained by the posture sensor, tire temperature obtained by the tire temperature sensor, vehicle data obtained by the vehicle controller, environmental data obtained by the weather station (needs to be explained) Figure 3The sensors shown are for illustrative purposes only; actual applications may include a greater variety or number of sensors. This data is then fed into the spatiotemporal alignment module for spatial and temporal alignment. After alignment, it passes through a cache queue and feature extraction to determine the track's curvature, slope, friction coefficient, and other parameters. This data is then fused into the data core to construct a corresponding digital track model.

[0106] Furthermore, based on the vehicle's dynamic parameters, the maximum acceleration and braking performance of the vehicle in the digital track model are calculated to obtain the dynamic constraint conditions.

[0107] Exemplary vehicle dynamics parameters include: vehicle mass, friction coefficient between tire and ground (μ), gravitational acceleration (g), vehicle power distribution, etc. Specifically, in practical applications, vehicle dynamics parameters may include vehicle mass (curb mass, total mass), wheelbase (distance between front and rear wheel centers), track (distance between left and right wheel centers), tire radius and width, friction coefficient between tire and ground (μ), vehicle center of gravity height, front and rear wheel load distribution, engine maximum output torque and power, torque distribution ratio (front and rear wheel power distribution), suspension stiffness and damping parameters, steering system parameters (such as steering ratio), braking system parameters (maximum braking force, distribution ratio), aerodynamic parameters (drag coefficient, lift coefficient, frontal area), etc. These parameters together determine the performance of the vehicle during dynamic processes such as acceleration, braking, and steering.

[0108] Calculate key dynamic constraints, such as maximum acceleration, which can include maximum dynamic acceleration and maximum lateral acceleration a y,max It is worth noting that the maximum lateral acceleration refers to the maximum lateral force that the tire can withstand when the vehicle is turning. The calculation formula is a y,max = μg, where μ is the coefficient of friction between the tire and the road, and g is the acceleration due to gravity. A larger value indicates a higher speed at which the vehicle can negotiate a curve.

[0109] The vehicle braking performance can include the critical braking distance d brake The critical braking distance refers to the shortest distance required for a vehicle to stop completely from a certain speed v during emergency braking. The calculation formula is d brake =v 2 / (2μg). This parameter can help determine how early to start braking on the track.

[0110] In addition, the dynamic constraints can also include the torque distribution ratio τ, which can represent the ratio of the power obtained by the front and rear wheels, τ = T front / T total , T front Front wheel power, T totalRear wheel power. This parameter affects the handling characteristics of the vehicle, such as the power distribution of front-wheel drive, rear-wheel drive, or four-wheel drive.

[0111] All the key parameters calculated above (such as a y,max , d brake , τ, etc.) can be embodied in the form of a table for subsequent reference and analysis. This table can contain more other parameters, such as maximum power speed, acceleration, etc.

[0112] In a feasible implementation, the shortest time for the vehicle to complete the track is taken as the simulation target, and the steps of simulating the driving process of the vehicle in the digital track model under the constraint of the dynamic constraint condition to generate the complete optimal driving strategy corresponding to the track include steps S210-S240:

[0113] Step S210, constructing a reward function with the shortest time for the vehicle to complete the track as the target;

[0114] Step S220, under the constraint of the dynamic constraint condition, simulating the driving process of the vehicle in the digital track model based on the current driving strategy to obtain an experience sample set, and adding the experience sample set to an experience replay pool, wherein the experience sample in the experience sample set is composed of a node vehicle state, a node vehicle action corresponding to the node vehicle state in the current driving strategy, a reward value of the node vehicle action determined by the reward function, and a next node vehicle state of the node vehicle state;

[0115] Step S230, collecting the experience samples in the experience replay pool, updating the network parameters of the strategy generation network, and generating a new current driving strategy through the updated strategy generation network;

[0116] Step S240, returning to execute the step of simulating the driving process of the vehicle in the digital track model based on the current driving strategy under the constraint of the dynamic constraint condition based on the new current driving strategy until the strategy generation network converges, and taking the driving strategy generated by the converged strategy generation network as the complete optimal driving strategy corresponding to the track.

[0117] It should be noted that in the present embodiment, the process of simulating the driving of the vehicle in the track is combined with the vehicle dynamics constraint, and the operation sequence of the accelerator, brake, and steering is iteratively optimized through a deep reinforcement learning algorithm (agent) to find the optimal driving strategy for the entire track.

[0118] For example, in order to facilitate the learning of the agent, the relevant information needs to be quantified. For example:

[0119] Constructing the state representation of the vehicle: s z ; constructing the action space of the vehicle: a; reward function: Rj =-T, where -T refers to a negative lap time. The shorter the lap time, the higher the reward.

[0120] Under the constraints of the above-mentioned dynamic constraints, the driving process of the vehicle in the digital track model is simulated based on the current driving strategy to obtain an experience sample set. Among them, the initial strategy can be set randomly, or the center line of the track can be directly used as the preliminary path to set the preset initial driving strategy. In the process of running a lap, an experience sample set consisting of a large number of experience samples can be generated. The experience sample consists of the node vehicle state, the node vehicle action corresponding to the node vehicle state in the current driving strategy, the reward value of the node vehicle action determined by the reward function, and the next node vehicle state of the node vehicle state. For example, the node vehicle state at time node t in the simulated running lap process can be expressed as s z，t , the current driving strategy is consistent with s z，t The corresponding node vehicle action can be expressed as a t , the reward value R of the node vehicle action determined by the reward function j，t , and the vehicle state at the next node is s t+1 That is, the composition of an experience sample is (s z，t , a t , R j，t , s z，t+1 ). The formed experience sample set is added to the experience replay pool for training and learning of the intelligent agent, that is, updating the parameters in the corresponding model of the intelligent agent. For example, each time training, a batch of experience samples can be collected in the experience replay pool, and then the policy generation network used to generate the policy can be trained using the loss function that matches the network model. That is, the model loss of the policy generation network is calculated by the loss function, and then the parameters in the policy generation network are updated by the model loss. After completing a training session, the updated policy generation network can be used to simulate the vehicle's driving process in the digital track model to generate a new current driving strategy. The new current driving strategy is used to generate new experience samples, that is, based on the new current driving strategy, the execution is returned to the step of simulating the vehicle's driving process in the digital track model based on the current driving strategy under the constraints of the dynamic constraints to obtain the experience sample set, until the policy generation network converges, and the driving strategy generated by the converged policy generation network can be used as the optimal driving strategy.

[0121] In practical applications, reinforcement learning algorithms can be DQN (Deep Q-Network) or PPO (Proximal Policy Optimization). The following examples illustrate this. The process of obtaining the optimal driving strategy is as follows:

[0122] Input: Track model (track space curve, curvature R(s), slope a(s), friction coefficient p(s), etc.), vehicle physical parameters (mass M, wheelbase L, maximum acceleration a max , maximum deceleration a min , maximum lateral acceleration a y,max , etc.).

[0123] Physical constraint modeling:

[0124] Calculate the maximum safe speed at each track position s:

[0125]

[0126] where R(s) is the radius of curvature, p(s) is the friction coefficient, and g is the acceleration due to gravity.

[0127] Path and speed field initialization:

[0128] Initialize the path (e.g., the track centerline).

[0129] Initialize the speed field V0(s) such that V0(s) ≤ v max (s).

[0130] Reinforcement learning optimization process:

[0131] State space x t : the state of the vehicle at time t (position s t , speed v t , acceleration a t , heading ψ t , etc.).

[0132] Action space u t : throttle δ throttle , brake δ brake , steering θ.

[0133] Reward function R t :

[0134] R j，t = -Δt

[0135] or R j，t = -1 for each time step, a terminal reward of +100, and a penalty of -100 for loss of control / deviation.

[0136] Objective: maximize the cumulative reward ∑ t R j，t , i.e., minimize the total lap time T.

[0137] Algorithm flow:

[0138] Initialize the agent policy π, usually as a random policy.

[0139] Construct a state space with vehicle status (such as position, speed, acceleration, heading, etc.) as input, and an action space with throttle, brake, steering, etc. as output.

[0140] The agent simulates multiple times in the virtual environment with the current strategy π, collecting the state-action-reward-next state sequence, that is, the experience sample.

[0141] Use algorithms such as DQN / PPO to update the strategy π to maximize the expected reward. The specific process is as follows:

[0142] In the virtual track environment, the agent performs multiple simulations according to the current strategy π, collecting the state-action-reward sequence (s z，t ,a t ,R j，t ,s t+1 ) and stored in the experience replay pool.

[0143] For the DQN algorithm, by sampling data from the experience replay pool, the Bellman equation is used to iteratively update the Q value network parameters so that Q(s,a) gradually approaches the optimal Q value. At each update, the following loss function is minimized:

[0144]

[0145] Where θ is the current Q network parameter, θ - is the target network parameter, γ is the discount factor, s z and a represent the state and action respectively. ′ z and a ′ Represent the state and action of the next node respectively.

[0146] The PPO algorithm uses a policy gradient method to optimize the policy network parameters to maximize the expected cumulative reward of the sampled trajectory. PPO also improves training stability by clipping the objective function to ensure that each policy update does not deviate too much.

[0147] During training, the reward function is usually set to negative lap time (i.e., the shorter the lap time, the higher the reward), and penalties are imposed for behaviors such as deviation from the track and loss of control, encouraging the agent to learn a driving strategy that is both safe and efficient.

[0148] After multiple rounds of iterative training, the strategy π gradually converges and ultimately outputs the optimal path and operation sequence that can maximize the expected reward (i.e., minimize lap time and improve driving performance).

[0149] After the strategy converges, the optimal path is output and velocity field V * (s).

[0150] Output:

[0151] Output the optimal path (sequence of spatial coordinate points) and recommended velocity field V * (s) (optimal speed at each position).

[0152] Parameter description and acquisition method:

[0153] R(s): Calculated by differentiating the three-dimensional model of the track.

[0154] μ(s): Fitted from sensor data such as tire temperature, road surface temperature, and humidity.

[0155] M, L: Vehicle factory parameters or actual measurements.

[0156] a max ,a min : Vehicle power / braking limit test or manufacturer's data.

[0157] Obtained through multimodal data fusion modeling, that is, the digital track model constructed.

[0158] In a feasible embodiment, before the step of constructing a digital track model simulating a track and the dynamic constraints of the vehicle, the method further includes steps S101 to S104:

[0159] Step S101: If a driving strategy of the same model as the vehicle exists in a preset strategy library corresponding to the track, an optimal driving strategy is obtained by matching the preset strategy library based on the vehicle state;

[0160] Step S102: If there is no driving strategy for the same model as the vehicle in the preset strategy library corresponding to the track, the similarity between the vehicle and the model corresponding to each driving strategy in the preset strategy library is calculated based on the vehicle parameters of the vehicle;

[0161] Step S103, when the maximum similarity among the similarities is less than or equal to a preset threshold, executing the step of constructing a digital track model of the simulated track and the dynamic constraints of the vehicle;

[0162] Step S104 : when the maximum similarity among the similarities is greater than a preset threshold, determining the optimal driving strategy of the vehicle based on the driving strategy of the vehicle model corresponding to the maximum similarity.

[0163] It should be noted that for new models of vehicles, the optimal driving strategy can be obtained by simulating laps. For old models or similar models that have already been simulated, they can be directly obtained. For example, when a vehicle starts the vehicle assisted driving function for the first time, it can first be matched with the vehicle model of each driving strategy in the preset strategy library corresponding to the track. If there is a driving strategy of the same model as the vehicle in the preset strategy library, the optimal driving strategy is obtained by matching in the preset strategy library based on the vehicle status of the vehicle, wherein the vehicle status may include the status of the vehicle itself, such as tire model, tire pressure, and vehicle power. The vehicle status may also include the temperature, humidity, and wind speed of the environment in which the vehicle is located. Obtaining the optimal driving strategy through vehicle status matching can ensure that the optimal driving strategy used for training is consistent with the current vehicle status, while ensuring the training effect.

[0164] Furthermore, if a driving strategy for the same vehicle model doesn't exist in the preset strategy library corresponding to the track, the similarity between the vehicle and the models corresponding to the various driving strategies in the preset strategy library is calculated based on the vehicle's parameters. For example, vehicle parameters can include length, weight, and power response. During the similarity matching process, these parameters are input into a neural network and converted into a "feature vector," a set of numbers describing the vehicle's characteristics. This vector is then compared with existing strategies. The preset strategy library contains multiple "baseline strategies," each corresponding to a specific vehicle model. The vehicle's feature vector is then compared with the characteristics of the models corresponding to these benchmark strategies to determine similarity.

[0165] If the maximum similarity among the obtained similarities is less than or equal to the preset threshold, it indicates that there is no existing strategy in the preset strategy library for the vehicle, and a new simulation is required to generate the optimal driving strategy. This involves executing the steps of constructing a digital track model of the simulated track and the vehicle's dynamic constraints to find the optimal driving strategy. It is worth noting that during the steps of constructing the digital track model of the simulated track and the vehicle's dynamic constraints and in the subsequent processes, the driving strategy corresponding to the maximum similarity can be used as the initial driving strategy for learning transfer, accelerating the search for the vehicle's optimal driving strategy.

[0166] Conversely, if the maximum similarity among the similarities is greater than a preset threshold, it indicates that the vehicle can use a ready-made driving strategy from the preset strategy library. The optimal driving strategy for the vehicle can be determined based on the driving strategy of the vehicle with the maximum similarity. For example, the driving strategy of the vehicle with the maximum similarity can be directly used as the optimal driving strategy, or the driving strategy of the vehicle with the maximum similarity can be fine-tuned to obtain the optimal driving strategy.

[0167] As reference Figure 4, is a flow chart of the strategy migration process in the application. For example, setting the target vehicle parameter c target =[L,M,τ,…], where L is the vehicle length, M is the vehicle mass, and τ is the torque distribution ratio. The target vehicle parameters are input into the corresponding neural network for feature encoding. In the feature encoding stage, the target vehicle parameters are mapped to a high-dimensional feature space for feature encoding to obtain a feature vector f target . Then calculate f target If the similarity is greater than 80% with the vehicle model characteristics corresponding to each strategy in the benchmark strategy library (i.e., the preset strategy), it is considered a high match, and the benchmark strategy library is loaded, and the matched strategy in the benchmark strategy library is fine-tuned, that is, the control parameters are fine-tuned to generate a migration strategy file. If it is less than or equal to 80%, it is considered a low match, and accordingly, the reinforcement learning module is started, virtual track training is performed, and a migration strategy file is generated. The migration strategy file is then deployed to the target vehicle ECU (Electronic Control Unit), and a verification test lap is performed, that is, whether the lap speed of the driving strategy deployed on the vehicle can meet the standard. If it meets the standard, the adaptation is completed, otherwise the feature encoding step is returned to repeat the above process.

[0168] In a feasible implementation manner, the vehicle assisted driving method further includes steps S01 and S02:

[0169] Step S01, monitoring the vehicle state of the vehicle, wherein the vehicle state includes the vehicle's own attribute state and the vehicle's environment state;

[0170] Step S02: After the change in the vehicle state exceeds a change threshold, the process returns to the step of matching the vehicle state with the preset strategy library to obtain the optimal driving strategy based on the vehicle state based on the changed vehicle state.

[0171] For example, after starting the vehicle assisted driving function, the vehicle computer will monitor the vehicle's vehicle status in addition to monitoring the vehicle's current driving state parameters. Among them, the vehicle status includes the vehicle's own attribute status and the vehicle's environmental status. The vehicle status that is more easily changed usually includes the tire pressure in the vehicle's own attribute status and the temperature, humidity and wind speed in the vehicle's environmental status. Similarly, the vehicle's vehicle status can be vectorized to facilitate the comprehensive calculation of the change in the vehicle status. After the change in the vehicle status exceeds the change threshold, the vehicle status after the change is returned to the step of matching the vehicle's vehicle status in the preset strategy library to obtain the optimal driving strategy.

[0172] Reference Figure 5, which is a schematic diagram of a scenario in this application. As shown in the figure, when the driver starts driving on the track, the tire pressure of the vehicle is monitored by the tire pressure sensor and suddenly begins to drop from 2.3 bar, and the rate of decrease is greater than 0.1 bar / s, that is, the change in the vehicle state exceeds the change threshold. Accordingly, the strategy optimizer strategy update is triggered, that is, the optimal driving strategy is re-matched, which causes the corresponding vehicle action to change and needs to be re-planned. For example, the braking point is recalculated (15 meters in advance). The path color coding is updated and output through the AR display unit (AR glasses), such as the path area within the current braking point range is passed through a red warning. The driver can adjust the braking force according to the warning, and finally when the tire pressure enters stability (that is, the stable value of 2.1 bar), the warning state is triggered and released.

[0173] It is understandable that, in this embodiment, the optimal driving strategy can be adjusted in real time according to changes in the vehicle state, thereby ensuring that the driving strategy used in the training is optimal throughout the entire process.

[0174] Reference below Figure 6 , which shows a schematic diagram of a controller structure suitable for implementing a vehicle according to an embodiment of the present application. The vehicle according to the embodiment of the present application may include, but is not limited to, mobile terminals such as computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as computers. Figure 6 The vehicle shown is merely an example and should not limit the functionality and scope of use of the embodiments of the present application.

[0175] like Figure 6As shown, the vehicle may include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes based on programs stored in a read-only memory (ROM) 1002 or programs loaded from a storage device 1003 into a random access memory (RAM) 1004. RAM 1004 also stores various programs and data required for vehicle operation. Processing device 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems may be connected to I / O interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speakers, and vibrator; storage device 1003 including, for example, a magnetic tape or hard disk; and communication device 1009. Communication device 1009 can allow the vehicle to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows a vehicle with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have instead.

[0176] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program comprising program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0177] The vehicle provided in this application, utilizing the vehicle-assisted driving method of the aforementioned embodiment, can address the technical problem that traditional training programs can only provide static references, exhibit significant lags, and lack real-time dynamic guidance, resulting in poor training effectiveness. Compared to the prior art, the beneficial effects of the vehicle provided in this application are the same as those of the vehicle-assisted driving method provided in the aforementioned embodiment, and the other technical features of the vehicle are the same as those disclosed in the aforementioned embodiment, and are not further described here.

[0178] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0179] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0180] The present application provides a computer-readable storage medium having computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the vehicle assisted driving method in the above-mentioned embodiment.

[0181] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0182] The computer-readable storage medium may be included in the vehicle, or may exist independently without being installed in the vehicle.

[0183] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by a vehicle, the vehicle:

[0184] Real-time monitoring of current driving state parameters of the vehicle;

[0185] Comparing the current driving state parameter with the optimal driving strategy corresponding to the track where the vehicle is located to obtain a state difference, wherein the state difference is the difference between the local driving strategy parameter corresponding to the current position of the vehicle in the optimal driving strategy and the current driving state parameter;

[0186] According to the state difference, vehicle correction prompt information is output.

[0187] Computer program code for performing the operations of the present application may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on the remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0188] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flow chart or block diagram can represent a module, program segment or a part of code, and the module, program segment or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order than that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented by a dedicated hardware-based system that performs the specified function or operation, or can be implemented by a combination of dedicated hardware and computer instructions.

[0189] The modules described in the embodiments of the present application may be implemented in software or hardware, wherein the name of a module does not necessarily limit the unit itself.

[0190] The readable storage medium provided in this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., a computer program) for executing the above-mentioned vehicle assisted driving method. It can solve the technical problem that traditional training programs can only provide static references, have obvious lags, lack real-time dynamic guidance value, and thus lead to poor training effects. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the vehicle assisted driving method provided in the above-mentioned embodiment, and will not be repeated here.

[0191] The present application also provides a computer program product, comprising a computer program, which implements the steps of the vehicle assisted driving method as described above when executed by a processor.

[0192] The computer program product provided in this application addresses the technical problem that traditional training programs only provide static references, exhibit significant lags, and lack real-time dynamic guidance, resulting in poor training effectiveness. Compared to the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the vehicle assisted driving method provided in the aforementioned embodiments, and are not further elaborated here.

[0193] The above description is only part of the embodiments of the present application and does not limit the patent scope of the present application. All equivalent structural transformations made by using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect application in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A vehicle assisted driving method, characterized in that: The vehicle assisted driving method comprises: Real-time monitoring of current driving state parameters of the vehicle; Comparing the current driving state parameter with the optimal driving strategy corresponding to the track where the vehicle is located to obtain a state difference, wherein the state difference is the difference between the local driving strategy parameter corresponding to the current position of the vehicle in the optimal driving strategy and the current driving state parameter; According to the state difference, vehicle correction prompt information is output.

2. The vehicle assisted driving method according to claim 1, wherein: Before the step of monitoring the current driving state parameters of the vehicle in real time, the method further includes: Constructing a digital track model simulating the track and the dynamic constraints of the vehicle; The shortest time for the vehicle to complete the track is set as a simulation goal. Under the constraints of the dynamic constraints, the driving process of the vehicle in the digital track model is simulated to generate an optimal driving strategy corresponding to the complete track.

3. The vehicle assisted driving method according to claim 2, wherein: The step of constructing a digital track model simulating the track and the dynamic constraints of the vehicle includes: Acquiring basic geometric data of the track, environmental data of the track, and a priori driving parameters, wherein the a priori driving parameters are collected by a priori vehicle traveling on the track; Building an initial model of the track based on the positioning data and posture data in the prior driving parameters; Extracting physical property features of the track from the geometric basic data and the environmental data, and loading the physical property features into the initial model to obtain the digital track model; The maximum acceleration and braking performance of the vehicle in the digital track model are calculated according to the dynamic parameters of the vehicle to obtain the dynamic constraint conditions.

4. The vehicle assisted driving method according to claim 2, wherein: The step of simulating the vehicle's driving process in the digital track model under the constraints of the dynamic constraints, taking the shortest time for the vehicle to complete the track as a simulation goal, and generating a complete optimal driving strategy corresponding to the track includes: Constructing a reward function with the goal of minimizing the time it takes for the vehicle to complete the track; Under the constraints of the dynamic constraints, simulating the driving process of the vehicle in the digital track model based on the current driving strategy to obtain an experience sample set, and adding the experience sample set to an experience replay pool, wherein the experience samples in the experience sample set consist of a node vehicle state, a node vehicle action corresponding to the node vehicle state in the current driving strategy, a reward value of the node vehicle action determined by the reward function, and a node vehicle state next to the node vehicle state; Collecting experience samples from the experience replay pool, updating network parameters of the strategy generation network, and generating a new current driving strategy through the updated strategy generation network; Based on the new current driving strategy, the step of simulating the driving process of the vehicle in the digital track model based on the current driving strategy under the constraints of the dynamic constraints to obtain an experience sample set is returned to be executed until the strategy generation network converges, and the driving strategy generated by the converged strategy generation network is used as the complete optimal driving strategy corresponding to the track.

5. The vehicle assisted driving method according to claim 2, wherein: Before the step of constructing a digital track model simulating the track and the dynamic constraints of the vehicle, the method further includes: If a driving strategy of the same model as that of the vehicle exists in a preset strategy library corresponding to the track, the optimal driving strategy is obtained by matching the driving strategies of the same model as that of the vehicle based on the vehicle state; If there is no driving strategy for the same model as that of the vehicle in the preset strategy library corresponding to the track, calculating the similarity between the vehicle and the models corresponding to each driving strategy in the preset strategy library based on the vehicle parameters of the vehicle; When the maximum similarity among the similarities is less than or equal to a preset threshold, executing the step of constructing a digital track model simulating the track and the dynamic constraint conditions of the vehicle; When a maximum similarity among the similarities is greater than a preset threshold, an optimal driving strategy for the vehicle is determined based on the driving strategy of the vehicle model corresponding to the maximum similarity.

6. The vehicle assisted driving method according to claim 5, characterized in that: The vehicle assisted driving method further includes: Monitoring the vehicle state of the vehicle, wherein the vehicle state includes the vehicle's own attribute state and the vehicle's environment state; After the change in the vehicle state of the vehicle exceeds a change threshold, the step of matching the vehicle state of the vehicle with the driving strategies of the same model as the vehicle to obtain the optimal driving strategy is returned based on the changed vehicle state.

7. The vehicle assisted driving method according to any one of claims 1 to 6, characterized in that: The virtual path in the optimal driving strategy and the driving parameters on the virtual path are projected into the field of view of the vehicle driver, the vehicle correction prompt information includes at least one of an oil brake operation prompt and a steering operation prompt, and the step of outputting the vehicle correction prompt information according to the state difference includes: If the state difference includes a speed difference, generating the oil brake operation prompt based on the speed difference and superimposing the oil brake operation prompt on the virtual path, wherein the oil brake operation prompt includes an accelerator action or a brake action, and an opening corresponding to the action; In a case where the state difference includes a position difference, generating the steering operation prompt based on the position difference, wherein the steering operation prompt includes a steering direction and a steering angle; By controlling the vibration of different positions on the steering wheel of the vehicle, the steering direction is prompted, and by controlling the vibration of the steering wheel to stop, it is prompted that the steering angle has been reached.

8. A vehicle, characterized in that: The vehicle includes: a processor, a memory, and a vehicle assisted driving program stored in the memory and runnable on the processor. When the vehicle assisted driving program is executed, the steps of the vehicle assisted driving method as described in any one of claims 1 to 7 are implemented.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a vehicle assisted driving program, which, when executed, implements the steps of the vehicle assisted driving method according to any one of claims 1 to 7.

10. A computer program product, characterized in that The computer program product includes a vehicle assisted driving program, which, when executed by a processor, implements the steps of the vehicle assisted driving method as described in any one of claims 1 to 7.

Images