A vehicle holographic perception model generation and management system based on the Internet of Things

Abstract

The present invention relates to the field of vehicle holographic perception model generation management, and specifically discloses a vehicle holographic perception model generation management system based on the Internet of Things. The present invention generates a vehicle holographic perception model by identifying conventional elements and unconventional elements in the vehicle driving environment and their detailed information, thereby improving the accuracy and reliability of the vehicle holographic perception model, so that the automatic driving system can make accurate, timely and safe decisions; the actual execution of the driving plan generated based on the vehicle holographic perception model is monitored, the execution effect of the vehicle driving plan is evaluated from the perspectives of safety, efficiency and riding experience, and the plan is stored, and the holographic perception model corresponding to the next driving environment of the vehicle is obtained, and the recommended reference driving plan for the next driving environment of the vehicle is selected from the similarity and execution effect of historical vehicle holographic perception models, thereby providing effective reference opinions for subsequent driving decision-making.

Description

Technical Field

[0001] The present invention relates to the field of vehicle holographic perception model generation management, and in particular to a vehicle holographic perception model generation management system based on the Internet of Things. Background Art

[0002] The vehicle holographic perception model is a system that integrates data from multiple sensors to create a detailed three-dimensional image of the surrounding environment. This model is designed to enable self-driving cars to perceive and understand their surroundings in an all-round, high-precision, and real-time manner. This perception is not limited to vision, but also includes the fusion of all detectable information to form a comprehensive environmental cognition model. The goal of holographic perception is to achieve 360-degree environmental monitoring without blind spots, and improve the safety and decision-making ability of self-driving vehicles.

[0003] The holographic perception model is the core of the autonomous driving system. It enables the vehicle to understand its environment and make safe and effective driving decisions. It is one of the key technologies to achieve fully autonomous driving.

[0004] There are some deficiencies in the existing generation and management methods of vehicle holographic perception models: 1. When generating vehicle holographic perception models, the existing methods do not collect enough dimensions of surrounding environment information. For example, they only collect information on common elements such as roads, pedestrians and vehicles in the surrounding environment, without considering the possibility of animals on the road. Moreover, since the comparison database used by vehicles to identify objects is limited, the existing methods do not take into account the possibility of unrecognizable objects in the surrounding environment. At the same time, the existing methods do not collect enough depth of surrounding environment information. They only collect the elements themselves in the surrounding environment, without further obtaining detailed information on the elements, such as type, location, speed, predicted motion trajectory, etc., which makes the generated vehicle holographic perception model low in accuracy and reliability, thereby reducing the safety and decision-making ability of autonomous driving vehicles.

[0005] 2. Different driving scenarios correspond to different driving plans. When formulating driving plans based on the vehicle holographic perception model, the existing methods do not monitor the actual execution of the driving plans generated based on the vehicle holographic perception model and then evaluate the effectiveness of the driving plans. As a result, when encountering similar driving scenarios in the future, it is impossible to provide a reference driving plan and provide effective reference opinions for subsequent driving decision-making. Summary of the invention

[0006] In view of the above problems, the present invention proposes a vehicle holographic perception model generation and management system based on the Internet of Things, which realizes the function of managing the generation of vehicle holographic perception models.

[0007] The technical solution adopted by the present invention to solve its technical problem is: the present invention provides a vehicle holographic perception model generation and management system based on the Internet of Things, including: a vehicle driving environment factor information acquisition module, a vehicle holographic perception model generation module, a vehicle preferred driving plan analysis module, a vehicle driving adjustment information acquisition feedback module, a vehicle preferred driving plan execution effect evaluation module, a vehicle holographic perception model and driving plan storage module, a vehicle reference driving plan recommendation module, and a database.

[0008] The vehicle holographic perception model generation module is respectively connected to the vehicle driving environment element information acquisition module and the vehicle preferred driving plan analysis module, the vehicle driving adjustment information acquisition feedback module is respectively connected to the vehicle preferred driving plan analysis module and the vehicle preferred driving plan execution effect evaluation module, the vehicle holographic perception model and driving plan storage module is connected to the vehicle preferred driving plan execution effect evaluation module, the vehicle reference driving plan recommendation module is connected to the vehicle holographic perception model and driving plan storage module, and the database is connected to the vehicle driving environment element information acquisition module.

[0009] Vehicle driving environment element information acquisition module: used to collect video data of the vehicle driving environment, and obtain information on conventional and unconventional elements of the vehicle driving environment, where conventional elements include traffic markings, traffic management facilities, other vehicles, pedestrians and animals, and unconventional elements include mobile roadblocks and stationary roadblocks.

[0010] Vehicle holographic perception model generation module: used to generate a vehicle's holographic perception model based on information about conventional and unconventional elements of the vehicle's driving environment.

[0011] Vehicle preferred driving plan analysis module: used to obtain the destination position of the vehicle, and combined with the vehicle's holographic perception model, obtain the driving plan that satisfies the road traffic rules, the principle of avoiding other vehicles, the principle of avoiding pedestrians, the principle of avoiding animals and the principle of avoiding roadblocks. It is recorded as the vehicle's preferred driving plan and bound to the vehicle's holographic perception model.

[0012] Vehicle driving adjustment information acquisition and feedback module: used to obtain the vehicle's driving adjustment information based on the vehicle's preferred driving plan and the vehicle's current driving information, and feed it back to the vehicle's control center.

[0013] Vehicle preferred driving plan execution effect evaluation module: used to obtain the safety factor, efficiency coefficient and riding experience coefficient of the vehicle's execution of the preferred driving plan, analyze the execution effect evaluation index of the vehicle's preferred driving plan, and bind it with the vehicle's preferred driving plan.

[0014] Vehicle holographic perception model and driving scheme storage module: used to store the vehicle's holographic perception model and the preferred driving scheme of the vehicle bound to it into the vehicle's data storage device.

[0015] Vehicle reference driving plan recommendation module: used to obtain video data of the vehicle's next driving environment, build a holographic perception model corresponding to the vehicle's next driving environment, analyze the recommended reference driving plan for the vehicle's next driving environment, and provide feedback.

[0016] Database: used to store sample image sets of traffic markings, traffic management facilities, vehicles, pedestrians and animals.

[0017] Compared with the prior art, the vehicle holographic perception model generation and management system based on the Internet of Things described in the present invention has the following beneficial effects: 1. The present invention generates a holographic perception model of the vehicle by identifying conventional elements such as traffic markings, traffic management facilities, other vehicles, pedestrians, animals, etc. in the vehicle driving environment, and unconventional elements such as mobile roadblocks and stationary roadblocks that cannot be identified by the vehicle identification object comparison database, and obtains detailed information on conventional and unconventional elements in the vehicle driving environment, thereby improving the accuracy and reliability of the generated vehicle holographic perception model, so that the autonomous driving system can make accurate, timely and safe decisions, thereby improving the safety and decision-making ability of autonomous driving vehicles.

[0018] 2. The present invention monitors the actual execution of the driving plan generated based on the vehicle holographic perception model, evaluates the execution effect of the vehicle driving plan from the perspectives of safety, efficiency and riding experience, stores the results, obtains the holographic perception model corresponding to the next driving environment of the vehicle, and selects the recommended reference driving plan for the next driving environment of the vehicle based on the similarity and execution effect of the historical vehicle holographic perception models, thereby providing effective reference opinions for subsequent driving decision-making and ensuring the safe driving of autonomous driving vehicles in complex traffic environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0020] Figure 1 This is a system module connection diagram of the present invention.

[0021] Figure 2 It is a schematic diagram of the vehicle holographic perception model architecture of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] See also Figure 1 and Figure 2 As shown, the present invention provides a vehicle holographic perception model generation and management system based on the Internet of Things, including a vehicle driving environment factor information acquisition module, a vehicle holographic perception model generation module, a vehicle preferred driving plan analysis module, a vehicle driving adjustment information acquisition feedback module, a vehicle preferred driving plan execution effect evaluation module, a vehicle holographic perception model and driving plan storage module, a vehicle reference driving plan recommendation module, and a database.

[0024] The vehicle holographic perception model generation module is respectively connected to the vehicle driving environment element information acquisition module and the vehicle preferred driving plan analysis module, the vehicle driving adjustment information acquisition feedback module is respectively connected to the vehicle preferred driving plan analysis module and the vehicle preferred driving plan execution effect evaluation module, the vehicle holographic perception model and driving plan storage module is connected to the vehicle preferred driving plan execution effect evaluation module, the vehicle reference driving plan recommendation module is connected to the vehicle holographic perception model and driving plan storage module, and the database is connected to the vehicle driving environment element information acquisition module.

[0025] The vehicle driving environment element information acquisition module is used to collect video data of the vehicle driving environment and obtain information on conventional elements and unconventional elements of the vehicle driving environment, wherein conventional elements include traffic markings, traffic management facilities, other vehicles, pedestrians and animals, and unconventional elements include mobile roadblocks and stationary roadblocks.

[0026] Furthermore, the specific working process of the vehicle driving environment element information acquisition module includes: E1: According to the preset principle, various visual sensors are evenly arranged around the vehicle body to collect environmental videos from all directions around the vehicle to obtain video data of the vehicle driving environment.

[0027] As a preferred solution, the visual sensor includes but is not limited to a high-definition camera and an infrared camera.

[0028] In a specific embodiment, there are a total of eight directions around the vehicle, including basic directions such as east, south, west, and north, and intermediate directions such as northeast, southeast, southwest, and northwest.

[0029] Based on the environmental videos in all directions around the vehicle, environmental images in all directions around the vehicle are obtained, and the real scene images of the vehicle's surrounding environment are obtained by splicing.

[0030] E2: Use image processing technology to obtain each grayscale value of the real scene image of the vehicle's surrounding environment, and compare it with the reference grayscale value range of the preset traffic marking image, analyze each traffic marking area of ​​the real scene image of the vehicle's surrounding environment, and further obtain the contour shape of each traffic marking area of ​​the real scene image of the vehicle's surrounding environment.

[0031] As a preferred solution, the traffic marking areas of the real-scene image of the vehicle's surrounding environment are analyzed. The specific method is: compare the grayscale values ​​of the real-scene image of the vehicle's surrounding environment with the preset reference grayscale value range of the traffic marking image. If a grayscale value of the real-scene image of the vehicle's surrounding environment belongs to the reference grayscale value range of the traffic marking image, then the grayscale value of the real-scene image of the vehicle's surrounding environment is recorded as a marked grayscale value, and the marked grayscale values ​​of the real-scene image of the vehicle's surrounding environment are counted. The areas where the marked grayscale values ​​of the real-scene image of the vehicle's surrounding environment are located are obtained and integrated to obtain the traffic marking areas of the real-scene image of the vehicle's surrounding environment.

[0032] As a preferred solution, when integrating the regions where the grayscale values ​​of the markers of the real scene image of the vehicle surroundings are located, the regions where the grayscale values ​​of the markers belong to the same connected region are spliced.

[0033] A sample image set of traffic markings stored in a database is extracted to obtain sample images of various types of traffic markings, and further the contour shapes of various types of traffic markings are obtained. The contour shapes of various traffic marking areas in the real scene image of the vehicle's surrounding environment are compared with the contour shapes of various types of traffic markings to obtain the types of various traffic markings in the vehicle's surrounding environment, and further the positions of various traffic markings in the vehicle's surrounding environment relative to the vehicle are obtained, which are recorded as the positions of various traffic markings in the vehicle's surrounding environment.

[0034] As a preferred solution, various types of traffic markings include lane dividing lines, guide arrows, pedestrian crossing lines, stop lines and other indication markings, as well as prohibition markings such as no overtaking lines, no lane change lines, no parking lines, and warning markings such as diamond markings in front of speed bumps, narrow road warning lines, and curve warning lines.

[0035] As a preferred solution, the position of the traffic marking relative to the vehicle can be represented by a combination of a direction angle and a distance.

[0036] As a preferred solution, the vehicle also carries instruments such as laser radar, ultrasonic sensors or infrared sensors to assist in measuring the distance between the vehicle and various elements of the vehicle's surrounding environment, and then obtain the position of various elements of the vehicle's surrounding environment relative to the vehicle.

[0037] E3: Obtain sample images of various types of traffic management facilities, and compare them with the real-life images of the vehicle's surroundings to identify the traffic management facilities and their types in the vehicle's surroundings, and further obtain the locations of the traffic management facilities in the vehicle's surroundings.

[0038] As a preferred solution, a sample image set of traffic management facilities stored in a database is extracted to obtain sample images of various types of traffic management facilities.

[0039] As a preferred option, using visual recognition technology, cameras can capture the appearance characteristics of traffic management facilities, such as color, shape and size, to distinguish different types of traffic management facilities.

[0040] As a preferred solution, various types of traffic management facilities include but are not limited to traffic lights and other signal lights, traffic signs such as construction signs, markings, guardrails, isolation belts, reflectors, traffic cones, etc.

[0041] Furthermore, the specific working process of the vehicle driving environment element information acquisition module also includes: obtaining sample images of various types of vehicles, and comparing them with real-life images of the vehicle's surroundings, identifying other vehicles and their types in the vehicle's surroundings, and further obtaining the positions of other vehicles in the vehicle's surroundings.

[0042] As a preferred solution, a sample image set of vehicles stored in a database is extracted to obtain sample images of various types of vehicles.

[0043] As a preferred option, using visual recognition technology, the camera can capture the vehicle's appearance features, such as color, outline and size, to distinguish different vehicle types.

[0044] In another specific embodiment, a laser radar carried by the vehicle is used to distinguish the type and location of other vehicles in the vehicle's surrounding environment.

[0045] Get the lighting status of other vehicles in the vehicle's surrounding environment.

[0046] As a preferred solution, based on the real-scene image of the vehicle's surroundings, images of other vehicles in the real-scene image of the vehicle's surroundings are obtained, and further the lighting states of other vehicles in the vehicle's surroundings are obtained.

[0047] As a preferred solution, the lighting status of other vehicles includes the lighting status of turn signals, brake lights, tail lights and other vehicle lights.

[0048] The video data of each other vehicle in the vehicle's surrounding environment is obtained to obtain the speed and movement trajectory of each other vehicle in the vehicle's surrounding environment.

[0049] As a preferred solution, video data of other vehicles in the surroundings of the vehicle are intercepted based on the video data of the vehicle's driving environment.

[0050] The lighting status and motion trajectory of each other vehicle in the vehicle's surrounding environment are substituted into a preset vehicle driving trajectory prediction model to obtain the predicted driving trajectory of each other vehicle in the vehicle's surrounding environment.

[0051] The method for obtaining the vehicle's driving trajectory prediction model is as follows: according to the vehicle's lighting status, the vehicle's turning direction, such as turning left, turning right or going straight, is obtained. Based on big data and machine learning algorithms, the vehicle's next possible movement trajectory is predicted through the vehicle's turning direction and a certain trajectory that has been traveled, and then the vehicle's driving trajectory prediction model is trained.

[0052] Furthermore, the specific working process of the vehicle driving environment element information acquisition module also includes: identifying various heat sources in the vehicle's surrounding environment through an infrared camera carried by the vehicle, and acquiring images of various heat sources in the vehicle's surrounding environment in combination with real-scene images of the vehicle's surrounding environment.

[0053] Sample images of various types of pedestrians and animals are obtained, and compared with the images of various heat sources in the vehicle's surrounding environment, to identify the pedestrians and their types, animals and their types in the vehicle's surrounding environment, and further obtain the positions of the pedestrians and animals in the vehicle's surrounding environment.

[0054] As a preferred solution, a set of sample images of pedestrians and a set of sample images of animals stored in a database are extracted to obtain sample images of various types of pedestrians and sample images of various types of animals.

[0055] In a specific embodiment, the types of pedestrians include children, middle-aged people, and elderly people.

[0056] As a preferred option, using visual recognition technology, the camera can capture the appearance characteristics of pedestrians, such as skin color, hair color, hairstyle, clothing, body shape and outline, so as to distinguish different types of pedestrians.

[0057] Obtain the body orientation of each pedestrian and animal in the vehicle's surrounding environment.

[0058] As a preferred solution, based on the real-scene image of the vehicle's surroundings, images of pedestrians and animals in the vehicle's surroundings are obtained, and the body orientations of pedestrians and animals in the vehicle's surroundings are further obtained.

[0059] The video data of each pedestrian and each animal in the environment around the vehicle are obtained to obtain the speed and activity trajectory of each pedestrian and each animal in the environment around the vehicle.

[0060] As a preferred solution, video data of pedestrians and animals in the environment surrounding the vehicle are intercepted based on the video data of the vehicle's driving environment.

[0061] The body orientation and activity trajectory of each pedestrian in the vehicle's surrounding environment are substituted into a preset pedestrian activity trajectory prediction model to obtain the predicted activity trajectory of each pedestrian in the vehicle's surrounding environment.

[0062] The pedestrian activity trajectory prediction model is obtained by obtaining the pedestrian's body orientation, such as east, south, west, and north, and based on big data and machine learning algorithms, predicting the pedestrian's next activity trajectory through the pedestrian's body orientation and a certain trajectory that has already been active, and then training the pedestrian's activity trajectory prediction model.

[0063] Similarly, based on the analysis method of the predicted activity trajectory of each pedestrian in the vehicle's surrounding environment, the predicted activity trajectory of each animal in the vehicle's surrounding environment is obtained, and the method of obtaining the animal's activity trajectory prediction model is the same as that of pedestrians.

[0064] Furthermore, the specific working process of the vehicle driving environment element information acquisition module also includes: acquiring elements in the real-scene image of the vehicle's surrounding environment that cannot be identified based on the sample image set stored in the database, and recording them as unconventional elements.

[0065] The speed of each unconventional element in the vehicle's surrounding environment is obtained, and each stationary unconventional element and each moving unconventional element in the vehicle's surrounding environment are obtained, which are respectively recorded as each stationary roadblock and each mobile roadblock in the vehicle's surrounding environment, and the speed of each mobile roadblock in the vehicle's surrounding environment is obtained.

[0066] As a preferred solution, based on the video data of the vehicle's driving environment, the video data of each unconventional element in the vehicle's surrounding environment is intercepted, and the speed of each unconventional element in the vehicle's surrounding environment is further obtained.

[0067] As a preferred solution, the speed of the stationary non-conventional element is zero.

[0068] The images of each stationary roadblock and each mobile roadblock in the surrounding environment of the vehicle are acquired, and the outlines and positions of each stationary roadblock and each mobile roadblock in the surrounding environment of the vehicle are obtained.

[0069] As a preferred solution, images of each stationary roadblock and each movable roadblock in the vehicle's surroundings are obtained based on the real-scene image of the vehicle's surroundings.

[0070] The video data of each mobile roadblock in the vehicle's surrounding environment is obtained to obtain the motion trajectory of each mobile roadblock in the vehicle's surrounding environment, which is then substituted into a preset motion trajectory prediction model of the mobile roadblock to obtain the predicted motion trajectory of each mobile roadblock in the vehicle's surrounding environment.

[0071] The motion trajectory prediction model of the mobile roadblock is obtained as follows: based on big data and machine learning algorithms, the possible next motion trajectory of the mobile roadblock is predicted through a section of the trajectory that the mobile roadblock has already moved, and then the motion trajectory prediction model of the mobile roadblock is trained.

[0072] As a preferred solution, video data of each mobile roadblock in the environment surrounding the vehicle is intercepted based on the video data of the vehicle's driving environment.

[0073] The vehicle holographic perception model generation module is used to generate a holographic perception model of the vehicle according to information of conventional elements and unconventional elements of the vehicle driving environment.

[0074] As a preferred solution, the specific working process of the vehicle holographic perception model generation module is as follows: a three-dimensional space coordinate system is established with the vehicle as the center, and the types and positions of traffic markings in the vehicle's surrounding environment, the types and positions of traffic management facilities, the types, positions, speeds and predicted driving trajectories of other vehicles, the types, positions, speeds and predicted activity trajectories of pedestrians, the types, positions, speeds and predicted activity trajectories of animals, the outlines and positions of stationary roadblocks, and the outlines, positions, speeds and predicted motion trajectories of mobile roadblocks are marked in the three-dimensional space coordinate system to generate a holographic perception model of the vehicle.

[0075] As a preferred solution, the vehicle's holographic perception model not only relies on on-board sensors such as cameras, lidar, millimeter-wave radar, and ultrasonic sensors, but may also combine real-time data from satellite navigation, map data, traffic information centers, weather forecast services, and vehicle-road cooperative systems.

[0076] It should be noted that the present invention generates a holographic perception model of the vehicle by identifying conventional elements such as traffic markings, traffic management facilities, other vehicles, pedestrians, animals, etc. in the vehicle driving environment, and unconventional elements such as mobile roadblocks and stationary roadblocks that cannot be identified by the vehicle identification object comparison database, and obtains detailed information on conventional and unconventional elements in the vehicle driving environment, thereby improving the accuracy and reliability of the generated vehicle holographic perception model, so that the autonomous driving system can make accurate, timely and safe decisions, thereby improving the safety and decision-making ability of autonomous driving vehicles.

[0077] The vehicle preferred driving plan analysis module is used to obtain the destination position of the vehicle, and in combination with the vehicle's holographic perception model, obtain a driving plan that satisfies the road traffic rules, the principle of avoiding other vehicles, the principle of avoiding pedestrians, the principle of avoiding animals and the principle of avoiding roadblocks. It is recorded as the vehicle's preferred driving plan and bound to the vehicle's holographic perception model.

[0078] As a preferred solution, the specific working process of the vehicle preferred driving plan analysis module is as follows: D1: obtaining the destination position of the vehicle.

[0079] D2: Obtain various driving plans that satisfy road traffic regulations based on the destination of the vehicle and the types and locations of traffic markings and traffic management facilities in the vehicle's surroundings.

[0080] D3: Obtain various driving plans that meet the principle of avoiding other vehicles based on the vehicle's destination location and the types, locations, speeds, and predicted driving trajectories of other vehicles in the vehicle's surrounding environment.

[0081] D4: Obtain various driving plans that meet the principle of avoiding pedestrians based on the destination location of the vehicle and the types, locations, speeds, and predicted activity trajectories of pedestrians in the vehicle's surrounding environment.

[0082] D5: Obtain various driving plans that meet the principle of avoiding animals based on the destination of the vehicle and the type, position, speed and predicted activity trajectory of each animal in the vehicle's surrounding environment.

[0083] D6: Based on the destination location of the vehicle and the contours and positions of each stationary roadblock in the vehicle's surrounding environment, as well as the contours, positions, speeds, and predicted motion trajectories of each mobile roadblock, various driving plans that satisfy the principle of avoiding roadblocks are obtained.

[0084] D7: Obtain a driving plan that satisfies the road traffic rules, the principle of avoiding other vehicles, the principle of avoiding pedestrians, the principle of avoiding animals, and the principle of avoiding roadblocks. Record it as the vehicle's preferred driving plan and bind it to the vehicle's holographic perception model.

[0085] As a preferred option, if there is more than one driving plan that satisfies the road traffic rules, the principle of avoiding other vehicles, the principle of avoiding pedestrians, the principle of avoiding animals and the principle of avoiding roadblocks, the one with the shortest distance shall be selected.

[0086] In a specific embodiment, when avoiding a stationary roadblock, the distance around the stationary roadblock is analyzed, and combined with the body width of the vehicle, a decision is made as to whether the vehicle should turn around or take a detour. If the vehicle takes a detour, the side from which the vehicle takes a detour is further analyzed.

[0087] In a specific embodiment, when avoiding a moving roadblock, the speed and predicted movement trajectory of the moving roadblock are analyzed to determine whether the moving roadblock is blocking the vehicle's progress. If so, a decision is made as to whether the vehicle or the moving roadblock should go first. If the vehicle goes first, the speed and route of the vehicle are analyzed. If the moving roadblock goes first, the vehicle continues to travel along the original route after the moving roadblock leaves.

[0088] How to avoid stationary and moving obstacles and generate optimal driving plans for vehicles involves multiple algorithms and technologies such as path planning algorithms, decision algorithms, control algorithms, fusion algorithms, and ethical decision algorithms. These technologies are integrated into the decision-making system of autonomous vehicles to help vehicles make safe driving decisions in complex road environments.

[0089] The vehicle driving adjustment information acquisition feedback module is used to obtain the vehicle's driving adjustment information according to the vehicle's preferred driving plan and in combination with the vehicle's current driving information, and feed it back to the vehicle's control center.

[0090] As a preferred solution, the specific working process of the vehicle driving adjustment information acquisition feedback module is: according to the vehicle's preferred driving plan, the vehicle's preferred driving route and speed are obtained, the vehicle's current driving information is obtained, the vehicle's current driving position and speed are obtained, the vehicle's driving position adjustment information and speed adjustment information are further analyzed, the vehicle's driving adjustment information is obtained, and the vehicle's driving adjustment information is fed back to the vehicle's control center.

[0091] As a preferred solution, the preferred speed of the vehicle is not necessarily a fixed value, but may also be a value that changes as the vehicle travels.

[0092] The vehicle preferred driving scheme execution effect evaluation module is used to obtain the safety factor, efficiency factor and riding experience factor of the vehicle's execution of the preferred driving scheme, analyze the execution effect evaluation index of the vehicle's preferred driving scheme, and bind it with the vehicle's preferred driving scheme.

[0093] Furthermore, the specific working process of the vehicle optimal driving scheme execution effect evaluation module includes: obtaining the number of traffic accidents and the number of near-accident events that occur during the vehicle's execution of the optimal driving scheme, and recording them as , by analyzing the formula Get the safety factor of the vehicle executing the optimal driving plan ,in represents the correction factor of the preset safety factor, They represent the impact factors corresponding to the preset number of traffic accidents per unit and number of near-accident events per unit, Respectively represent the weights of preset traffic accidents and near-accident events, , .

[0094] As a preferred option, traffic accidents include minor collisions, scrapes, etc.

[0095] As a preferred option, near accident events refer to situations that do not cause damage but are potentially dangerous, such as emergency braking, evasive actions, etc.

[0096] Furthermore, the specific working process of the vehicle preferred driving scheme execution effect evaluation module also includes: obtaining the driving time and energy consumption of the vehicle executing the preferred driving scheme, and recording them as , obtain the driving distance of the vehicle's preferred driving plan, analyze the reference driving time and reference energy consumption of the vehicle's preferred driving plan, and record them as , by analyzing the formula Get the efficiency coefficient of the vehicle executing the optimal driving plan ,in Indicates the correction factor of the preset efficiency coefficient, represents a natural constant, Respectively represent the weights of the preset driving time and energy consumption, .

[0097] As a preferred solution, the energy consumption of the vehicle includes but is not limited to: fuel consumption and electricity consumption.

[0098] Furthermore, the specific working process of the vehicle preferred driving scheme execution effect evaluation module also includes: obtaining the number of sudden acceleration, sudden deceleration and sudden braking of the vehicle during the execution of the preferred driving scheme, and recording them as , by analyzing the formula Get the riding experience coefficient of the vehicle executing the optimal driving plan ,in Indicates the correction factor of the preset riding experience coefficient, They respectively represent the influencing factors corresponding to the preset unit times of rapid acceleration, unit times of rapid deceleration and unit times of rapid braking.

[0099] Furthermore, the specific working process of the vehicle optimal driving scheme execution effect evaluation module also includes: by analyzing the formula Get the execution effect evaluation index of the vehicle's optimal driving plan ,in They represent the weights of the preset safety factor, efficiency factor and riding experience factor respectively. , bind it to the vehicle's preferred driving plan, and store it in the vehicle's data storage device.

[0100] The vehicle holographic perception model and driving scheme storage module is used to store the vehicle's holographic perception model and the preferred driving scheme of the vehicle bound thereto into the vehicle's data storage device.

[0101] The vehicle reference driving plan recommendation module is used to obtain video data of the vehicle's next driving environment, construct a holographic perception model corresponding to the vehicle's next driving environment, analyze the recommended reference driving plan for the vehicle's next driving environment, and provide feedback.

[0102] Furthermore, the specific working process of the vehicle reference driving plan recommendation module is: obtaining video data of the vehicle's next driving environment and constructing a holographic perception model corresponding to the vehicle's next driving environment.

[0103] The execution effect evaluation indexes of the various holographic perception models of the vehicle and the preferred driving schemes bound thereto stored in the data storage device of the vehicle are retrieved and recorded as the execution effect evaluation indexes of the various historical holographic perception models and the preferred driving schemes bound thereto.

[0104] Obtain the similarity between each historical holographic perception model and the holographic perception model corresponding to the next driving environment of the vehicle.

[0105] The weighted average of the similarity between each historical holographic perception model and the holographic perception model corresponding to the next driving environment of the vehicle and the execution effect evaluation index of the preferred driving scheme bound to each historical holographic perception model is calculated to obtain the recommendation coefficient of each historical holographic perception model.

[0106] As a preferred solution, the weights of the similarity and the execution effect evaluation index are set values, and the cumulative sum is 1.

[0107] The bound optimal driving plan of the historical holographic perception model corresponding to the maximum recommendation coefficient is used as the recommended reference driving plan for the next driving environment of the vehicle.

[0108] It should be noted that the present invention monitors the actual execution of the driving plan generated based on the vehicle holographic perception model, evaluates the execution effect of the vehicle driving plan from the perspectives of safety, efficiency and riding experience, stores the results, obtains the holographic perception model corresponding to the next driving environment of the vehicle, and selects the recommended reference driving plan for the next driving environment of the vehicle based on the similarity and execution effect of the historical vehicle holographic perception models, thereby providing effective reference opinions for subsequent driving decision-making and ensuring the safe driving of autonomous driving vehicles in complex traffic environments.

[0109] The database is used to store sample image sets of traffic markings, traffic management facilities, vehicles, pedestrians and animals.

[0110] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they shall all fall within the protection scope of the present invention.

Claims

1. A vehicle holographic perception model generation and management system based on the Internet of Things, characterized in that: include: Vehicle driving environment element information acquisition module: used to collect video data of the vehicle driving environment, and obtain information on conventional elements and unconventional elements of the vehicle driving environment, where conventional elements include traffic markings, traffic management facilities, other vehicles, pedestrians and animals, and unconventional elements include mobile roadblocks and stationary roadblocks; Vehicle holographic perception model generation module: used to generate a vehicle holographic perception model based on information on conventional and unconventional elements of the vehicle's driving environment; Vehicle optimal driving plan analysis module: used to obtain the destination location of the vehicle, and in combination with the vehicle's holographic perception model, obtain a driving plan that satisfies the road traffic rules, the principle of avoiding other vehicles, the principle of avoiding pedestrians, the principle of avoiding animals, and the principle of avoiding roadblocks. It is recorded as the vehicle's optimal driving plan and bound to the vehicle's holographic perception model; Vehicle driving adjustment information acquisition and feedback module: used to obtain the vehicle's driving adjustment information according to the vehicle's preferred driving plan and in combination with the vehicle's current driving information, and feed it back to the vehicle's control center; Vehicle preferred driving plan execution effect evaluation module: used to obtain the safety factor, efficiency factor and riding experience factor of the vehicle's preferred driving plan, analyze the execution effect evaluation index of the vehicle's preferred driving plan, and bind it with the vehicle's preferred driving plan; Vehicle holographic perception model and driving scheme storage module: used to store the vehicle's holographic perception model and the vehicle's bound preferred driving scheme into the vehicle's data storage device; Vehicle reference driving plan recommendation module: used to obtain video data of the vehicle's next driving environment, build a holographic perception model corresponding to the vehicle's next driving environment, analyze the recommended reference driving plan for the vehicle's next driving environment, and provide feedback; Database: used to store sample image sets of traffic markings, traffic management facilities, vehicles, pedestrians and animals; The specific working process of the vehicle optimal driving scheme execution effect evaluation module includes: Obtain the number of traffic accidents and near-accident events that occur during the vehicle's execution of the optimal driving plan, and record them as The near accident event refers to a situation that does not cause damage but has potential danger. Get the safety factor of the vehicle executing the optimal driving plan ,in represents the correction factor of the preset safety factor, They represent the impact factors corresponding to the preset number of traffic accidents per unit and number of near-accident events per unit, Respectively represent the weights of preset traffic accidents and near-accident events, , ; The specific working process of the vehicle optimal driving scheme execution effect evaluation module also includes: Obtain the driving time and energy consumption of the vehicle executing the optimal driving plan, and record them as , obtain the driving distance of the vehicle's preferred driving plan, analyze the reference driving time and reference energy consumption of the vehicle's preferred driving plan, and record them as , by analyzing the formula Get the efficiency coefficient of the vehicle executing the optimal driving plan ,in Indicates the correction factor of the preset efficiency coefficient, represents a natural constant, Respectively represent the weights of the preset driving time and energy consumption, ; The specific working process of the vehicle optimal driving scheme execution effect evaluation module also includes: Obtain the number of sudden acceleration, deceleration and braking of the vehicle during the execution of the optimal driving plan, and record them as , by analyzing the formula Get the riding experience coefficient of the vehicle executing the optimal driving plan ,in Indicates the correction factor of the preset riding experience coefficient, Respectively represent the influence factors corresponding to the preset unit number of rapid acceleration, unit number of rapid deceleration and unit number of rapid braking; The specific working process of the vehicle optimal driving scheme execution effect evaluation module also includes: By analyzing the formula Get the execution effect evaluation index of the vehicle's optimal driving plan ,in They represent the weights of the preset safety factor, efficiency factor and riding experience factor respectively. , bind it with the vehicle's preferred driving plan, and store it in the vehicle's data storage device; The specific working process of the vehicle reference driving plan recommendation module is as follows: Obtain video data of the next driving environment of the vehicle and construct a holographic perception model corresponding to the next driving environment of the vehicle; Retrieving the execution effect evaluation indexes of each holographic perception model of the vehicle and the preferred driving schemes bound thereto stored in the data storage device of the vehicle, and recording them as the execution effect evaluation indexes of each historical holographic perception model and the preferred driving schemes bound thereto; Obtaining the similarity between each historical holographic perception model and the holographic perception model corresponding to the next driving environment of the vehicle; The weighted average of the similarity between each historical holographic perception model and the holographic perception model corresponding to the next driving environment of the vehicle and the execution effect evaluation index of each historical holographic perception model bound to the optimal driving plan is calculated to obtain the recommendation coefficient of each historical holographic perception model; The bound optimal driving plan of the historical holographic perception model corresponding to the maximum recommendation coefficient is used as the recommended reference driving plan for the next driving environment of the vehicle.

2. The vehicle holographic perception model generation and management system based on the Internet of Things according to claim 1 is characterized by: The specific working process of the vehicle driving environment element information acquisition module includes: E1: According to the preset principle, various visual sensors are evenly arranged around the vehicle body to collect environmental videos from all directions around the vehicle and obtain video data of the vehicle's driving environment; Based on the environmental videos in all directions around the vehicle, environmental images in all directions around the vehicle are obtained, and the real scene images of the environment around the vehicle are obtained by splicing; E2: using image processing technology to obtain each grayscale value of the real scene image of the vehicle's surrounding environment, and compare it with the preset reference grayscale value range of the traffic marking image, analyzing each traffic marking area of ​​the real scene image of the vehicle's surrounding environment, and further obtaining the contour shape of each traffic marking area of ​​the real scene image of the vehicle's surrounding environment; Extracting a sample image set of traffic markings stored in a database, obtaining sample images of various types of traffic markings, further obtaining contour shapes of various types of traffic markings, comparing contour shapes of various traffic marking areas in the real scene image of the vehicle surrounding environment with contour shapes of various types of traffic markings, obtaining types of various traffic markings in the vehicle surrounding environment, further obtaining positions of various traffic markings in the vehicle surrounding environment relative to the vehicle, and recording them as positions of various traffic markings in the vehicle surrounding environment; E3: Obtain sample images of various types of traffic management facilities, and compare them with the real-life images of the vehicle's surroundings to identify the traffic management facilities and their types in the vehicle's surroundings, and further obtain the locations of the traffic management facilities in the vehicle's surroundings.

3. The vehicle holographic perception model generation and management system based on the Internet of Things according to claim 2 is characterized by: The specific working process of the vehicle driving environment element information acquisition module also includes: Obtain sample images of various types of vehicles, and compare them with real-life images of the vehicle's surroundings, identify other vehicles and their types in the vehicle's surroundings, and further obtain the positions of other vehicles in the vehicle's surroundings; Obtain the lighting status of other vehicles in the vehicle's surrounding environment; Acquire video data of other vehicles in the environment around the vehicle, and obtain the speed and movement trajectory of other vehicles in the environment around the vehicle; The lighting status and motion trajectory of each other vehicle in the vehicle's surrounding environment are substituted into a preset vehicle driving trajectory prediction model to obtain the predicted driving trajectory of each other vehicle in the vehicle's surrounding environment.

4. The vehicle holographic perception model generation and management system based on the Internet of Things according to claim 3 is characterized by: The specific working process of the vehicle driving environment element information acquisition module also includes: The infrared camera carried by the vehicle is used to identify various heat sources in the environment around the vehicle, and the images of various heat sources in the environment around the vehicle are obtained by combining the real-life images of the environment around the vehicle; Obtain sample images of various types of pedestrians and animals, and compare them with the images of various heat sources in the vehicle's surrounding environment, identify the pedestrians and their types, animals and their types in the vehicle's surrounding environment, and further obtain the positions of the pedestrians and animals in the vehicle's surrounding environment; Obtain the body orientation of each pedestrian and animal in the vehicle's surrounding environment; Obtain video data of pedestrians and animals in the vehicle's surrounding environment, and obtain the speed and activity trajectory of pedestrians and animals in the vehicle's surrounding environment; Substituting the body orientation and activity trajectory of each pedestrian in the vehicle surrounding environment into a preset pedestrian activity trajectory prediction model to obtain the predicted activity trajectory of each pedestrian in the vehicle surrounding environment; Similarly, according to the analysis method of the predicted activity trajectory of each pedestrian in the environment around the vehicle, the predicted activity trajectory of each animal in the environment around the vehicle is obtained.

5. The vehicle holographic perception model generation and management system based on the Internet of Things according to claim 4 is characterized by: The specific working process of the vehicle driving environment element information acquisition module also includes: Obtaining elements in the real scene image of the vehicle's surrounding environment that cannot be identified based on the sample image set stored in the database, and recording them as unconventional elements; Obtaining the speed of each unconventional element in the environment around the vehicle, obtaining each stationary unconventional element and each moving unconventional element in the environment around the vehicle, recording them as each stationary roadblock and each moving roadblock in the environment around the vehicle, and obtaining the speed of each moving roadblock in the environment around the vehicle; Acquire images of each stationary roadblock and each mobile roadblock in the environment around the vehicle, and obtain contours and positions of each stationary roadblock and each mobile roadblock in the environment around the vehicle; The video data of each mobile roadblock in the vehicle's surrounding environment is obtained to obtain the motion trajectory of each mobile roadblock in the vehicle's surrounding environment, which is then substituted into a preset motion trajectory prediction model of the mobile roadblock to obtain the predicted motion trajectory of each mobile roadblock in the vehicle's surrounding environment.

Images