Automatic driving system and method of two-wheeled electric vehicle

By integrating navigation, sensors, computing and power systems on two-wheeled electric vehicles, the global and local path planning of the autonomous driving system is achieved, and the safety and efficiency problems caused by the low quality of the drivers of two-wheeled electric vehicles are solved, and automatic Driving stability and safety.

CN119018273AInactive Publication Date: 2024-11-26GREATER BAY AREA INST FOR INNOVATION HUNAN UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411188457.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-28
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

During driving, two-wheeled electric vehicles are not very good at drivers, are prone to traffic accidents, and insufficient application of autonomous driving technology, resulting in safety and efficiency problems.

Method used

A two-wheeled electric vehicle autonomous driving system including navigation system, sensor system, computing platform, power system and self-balancing system was designed. Data is collected in real time through sensors such as GPS, lidar, camera, acceleration sensor and gyroscope, and SLAM is used to collect data in real time. The algorithm builds a road environment model and carries out global and local path planning to ensure that the electric vehicle drives stably during autonomous driving.

Benefits of technology

It improves the driving safety and driving efficiency of two-wheeled electric vehicles, reduces the probability of traffic accidents, meets the self-balancing needs of two-wheeled electric vehicles, and obtains services through the Internet, improving the safety and reliability of autonomous driving.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119018273A_ABST
    Figure CN119018273A_ABST
Patent Text Reader

Abstract

The invention belongs to the technical field of automatic driving of two-wheeled electric vehicles, and particularly relates to an automatic driving system and method of a two-wheeled electric vehicle, and the automatic driving system comprises a navigation system, a sensor system, a computing platform, a power system and a self-balancing system. The navigation system is composed of a GPS module and a wireless communication module. The sensor system is composed of a laser radar, a camera, an acceleration sensor and a gyroscope sensor. The computing platform is composed of a target recognition module, a three-dimensional reconstruction module, a path planning module and a control module. The power system is composed of a torque motor, a motor and a brake. The self-balancing system is composed of a deflection motor, a flywheel, a supporting foot stand and an auxiliary wheel.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of automatic driving of two-wheeled electric vehicles, and in particular relates to an automatic driving system and method for a two-wheeled electric vehicle. Background Art

[0002] With the development of new energy technology and the continuous rise in oil prices, electric vehicles have gradually replaced motorcycles and become people's main short-distance transportation tool. In view of the frequent traffic accidents involving electric vehicles, the reasons for developing electric vehicle autonomous driving are listed as follows:

[0003] Since driving an electric vehicle does not require any qualification assessment, the average quality of electric vehicle drivers is not high. It has become common for electric vehicles to change lanes and run red lights at will on the road, which often leads to traffic accidents. The automatic driving system can strictly abide by the traffic rules set by the program and quickly identify the driving risks hidden in complex road conditions (such as the blind spots of large vehicles), thereby ensuring driving efficiency while avoiding serious traffic accidents.

[0004] The human body may experience adverse reactions such as fatigue and dizziness during driving, which may endanger driving safety. However, the failure probability of two-wheeled electric vehicles and automatic driving systems that have undergone strict inspections is low, which can improve driving safety.

[0005] Due to the rapid development of chip performance and artificial intelligence technology, automobile autonomous driving has been highly developed. However, the autonomous driving of two-wheeled electric vehicles, which are easy to charge, flexible to control, and have a larger market share, has received little attention. Therefore, it is necessary and feasible to migrate existing automobile autonomous driving technologies and solutions to two-wheeled electric vehicles. Summary of the invention

[0006] The purpose of the present invention is to provide an automatic driving system and method for a two-wheeled electric vehicle, so as to solve the problems existing in the background technology.

[0007] In order to achieve the above technical objectives, the technical solution adopted by the present invention is as follows:

[0008] An automatic driving system for a two-wheeled electric vehicle, comprising a navigation system, a sensor system, a computing platform, a power system, and a self-balancing system;

[0009] The navigation system is composed of a GPS module and a wireless communication module;

[0010] The sensor system is composed of a laser radar, a camera, an acceleration sensor and a gyroscope sensor;

[0011] The computing platform is composed of a target recognition module, a three-dimensional reconstruction module, a path planning module and a control module;

[0012] The power system is composed of a torque motor, an electric motor and a brake;

[0013] The self-balancing system consists of a deflection motor, a flywheel, a supporting tripod and auxiliary wheels.

[0014] The automatic driving method comprises the following steps:

[0015] Step 1: Global route planning

[0016] The global route planning is performed by locating the vehicle's own position using GPS and using a wireless communication module to request navigation services from a cloud server. The cloud server plans the route of the two-wheeled electric vehicle in a city map based on satellite positioning, and calculates the shortest and closest route based on the road traffic, speed, environment and weather conditions of the Internet of Vehicles system, and determines and returns the driving range and direction of the two-wheeled electric vehicle.

[0017] Step 2: Real-time collection of road information

[0018] The laser radar obtains real-time information about obstacles on the road, and the camera obtains real-time visual information of the road scene. At the same time, the acceleration sensor and gyroscope sensor record the posture changes of the electric vehicle, including acceleration and angular velocity, to provide auxiliary information for positioning;

[0019] After acquiring and processing the sensor data, the SLAM algorithm begins to build a road environment model to plan the electric vehicle's driving path in the future, and builds a three-dimensional network model. The three-dimensional grid model is projected onto a two-dimensional plane to form a two-dimensional local road map, which identifies the movable area, obstacles and other road participant information. The SLAM algorithm realizes real-time positioning of the electric vehicle, matches the sensor data with the constructed map, and determines the location of the device in the map;

[0020] Step 3: Local path planning

[0021] Based on the constructed three-dimensional map model, the path planning system dynamically plans the specific operation path of the two-wheeled electric vehicle;

[0022] Step 4: Steering Control

[0023] According to the determined path and the current state of the vehicle including speed and heading angle, the control system of the present invention ensures that the two-wheeled electric vehicle moves along the planned path by controlling the front wheel steering angle.

[0024] Step 5: Self-balancing electric vehicle

[0025] The deflection motor, flywheel, support tripod and auxiliary wheels cooperate with the gyroscope to ensure the stability of the electric vehicle during turning, driving, driving at the lowest speed, driving at the highest speed and waiting for parking during the automatic driving process.

[0026] In global route planning, a road graph model is constructed based on satellite images, city maps and other data. In the graph model, buildings, roads, public facilities, etc. in the city are regarded as nodes, and each node has a unique identifier and corresponding attributes.

[0027] Identifiers and corresponding attributes include name, position, height and type, and define the relationships between nodes, such as adjacent relationships, connection relationships, etc. These relationships can be represented in the graph model in the form of edges.

[0028] The navigation system uses the wireless communication module to request the navigation map from the cloud server, and obtains the basic information of the electric vehicle's position and direction in the map in real time through the GPS module.

[0029] Various traffic signs and signal light status can be identified through the camera.

[0030] The local path planning has the following limitations: the driving path should be limited to the road range planned in step 1; the next path point should be within a fan of a given angle centered on the current vehicle head direction; any path point cannot be within a safe distance from a stationary obstacle; when reaching any path point, it cannot overlap with the predicted path of a moving obstacle in the current environment; the moving path needs to meet the smoothness requirement, and the product of the curvature and the speed cannot be greater than the limit value to ensure the balance of the two-wheeled electric vehicle.

[0031] When calculating the route in global route planning, 3-5 candidate routes with the shortest time and closest distance are calculated. Road information such as road flow and traffic speed are collected based on the Internet of Vehicles system, and real-time weather information is obtained from local meteorological stations. The above information is analyzed to select the best route.

[0032] The automatic driving system for two-wheeled electric vehicles described in the present invention has the following characteristics: there are great differences between two-wheeled electric vehicles and automobiles in weight, size, and power, which leads to significant differences in the automatic driving schemes of the two; the light weight and unstable supporting structure of two-wheeled electric vehicles lead to poor self-balancing and the risk of tipping over, which requires that the two-wheeled electric vehicles need to include self-balancing devices and systems, and adopt different balancing strategies at different stages of the operation of the electric vehicles; the size of two-wheeled electric vehicles is much smaller than that of automobiles, and can only carry lightweight sensors and intelligent computing devices, so the automatic driving of two-wheeled electric vehicles needs to obtain services through the Internet; because two-wheeled electric vehicles do not have protective equipment for drivers and passengers, the safety requirements of two-wheeled electric vehicles during the automatic driving process are higher. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The present invention is further illustrated by means of the non-limiting examples given in the accompanying drawings.

[0034] Figure 1 The process of the present invention is shown in FIG. Figure 1 ;

[0035] Figure 2 The process of the present invention is shown in FIG. Figure 2 . DETAILED DESCRIPTION

[0036] In order to enable those skilled in the art to better understand the present invention, the technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0037] like Figure 1-2 As shown, an automatic driving system for a two-wheeled electric vehicle of the present invention includes a navigation system, a sensor system, a computing platform, a power system, and a self-balancing system;

[0038] The navigation system is composed of a GPS module and a wireless communication module;

[0039] The sensor system is composed of a laser radar, a camera, an acceleration sensor and a gyroscope sensor;

[0040] The computing platform is composed of a target recognition module, a three-dimensional reconstruction module, a path planning module and a control module;

[0041] The power system is composed of a torque motor, an electric motor and a brake;

[0042] The self-balancing system consists of a deflection motor, a flywheel, a supporting tripod and auxiliary wheels.

[0043] The automatic driving method comprises the following steps:

[0044] Step 1: Global route planning

[0045] The global route planning is performed by locating the position of the vehicle itself with GPS, and using the wireless communication module to request navigation services from the cloud server. The cloud server plans the route of the two-wheeled electric vehicle in the city map based on satellite positioning, and calculates the shortest and closest route based on the road traffic, speed, environment and weather conditions of the Internet of Vehicles system, and determines and returns the driving range and driving direction of the two-wheeled electric vehicle.

[0046] Step 2: Real-time collection of road information

[0047] The laser radar obtains real-time information about obstacles on the road, and the camera obtains real-time visual information of the road scene. At the same time, the acceleration sensor and gyroscope sensor record the posture changes of the electric vehicle, including acceleration and angular velocity, to provide auxiliary information for positioning;

[0048] After acquiring and processing the sensor data, the SLAM algorithm begins to build a road environment model to plan the electric vehicle's driving path in the future, and builds a three-dimensional network model. The three-dimensional grid model is projected onto a two-dimensional plane to form a two-dimensional local road map, which identifies the movable area, obstacles and other road participant information. The SLAM algorithm realizes real-time positioning of the electric vehicle, matches the sensor data with the constructed map, and determines the location of the device in the map;

[0049] Step 3: Local path planning

[0050] Based on the constructed three-dimensional map model, the path planning system dynamically plans the specific operation path of the two-wheeled electric vehicle;

[0051] Step 4: Steering Control

[0052] According to the determined path and the current state of the vehicle including speed and heading angle, the control system of the present invention ensures that the two-wheeled electric vehicle moves along the planned path by controlling the front wheel steering angle;

[0053] Step 5: Self-balancing electric vehicle

[0054] The deflection motor, flywheel, support tripod and auxiliary wheel cooperate with the gyroscope to ensure the stability of the electric vehicle during turning, driving, driving at the lowest speed, driving at the highest speed and waiting for parking during the automatic driving process;

[0055] In the implementation of the technical features of the present invention, the SLAM technology is used to construct a map for the electric vehicle's own positioning and driving. The functions of using this technology are: 1. Identify and extract features of targets in the surrounding environment through sensors; 2. Reconstruct feature landmarks in three dimensions through the obtained information; 3. Continuously obtain new feature surfaces in the surrounding area during movement, and correct the map model in the mind; 4. Determine its own information based on the obtained feature landmarks and the obtained map model; 5. Determine whether the positions passed by the path are repeated based on the comparison and identification of landmarks, and self-correct to select the optimal route.

[0056] The present invention obtains the required information from satellite images and image data obtained in real time. Therefore, it is necessary to correspond and combine the pixel points obtained in the image data with the three-dimensional world coordinates to determine the accuracy of the real-time three-dimensional environment map. Therefore, it is necessary to establish an imaging model and calibrate the camera.

[0057] In the process of collecting real-time road information, a four-line laser radar is used to measure the three-dimensional point cloud of road obstacles to provide high-precision depth information and provide an information basis for electric vehicle automatic driving obstacle avoidance. The radar transmitter emits a laser beam. After the laser beam encounters an object, it is diffusely reflected and returned to the laser receiver. The radar module multiplies the time interval between sending and receiving signals by the speed of light and then divides it by 2 to calculate the distance between the transmitter and the object. At the same time, the four-line laser radar polls the four laser transmitters. After a polling cycle, a frame of laser point cloud data is obtained. The four point cloud data can form surface information, so that the height information of the obstacle can be obtained. The distance information of the obstacle can be obtained according to the point cloud coordinates of a single frame. According to the coordinates of the point cloud of multiple frames, the distance information is differentiated to obtain the speed information of the obstacle.

[0058] When using the sensor system to identify other road participants such as cars, pedestrians, traffic signs such as traffic lights, and electric vehicles for acceleration and angular velocity information collection, all types of collected information are pre-processed, including: denoising, data synchronization, image information processing, point cloud model registration and point cloud data fusion;

[0059] The purpose of denoising is to remove noise from the original data by filtering and other methods;

[0060] The purpose of data synchronization is to synchronize the time of data from different sensors to ensure data consistency;

[0061] The purpose of image information processing is to use the SFM algorithm to reconstruct the three-dimensional point cloud of the image information collected by the camera to obtain a three-dimensional point cloud model with pedestrian and vehicle information;

[0062] The purpose of point cloud model registration is to extract key feature points, such as corner points and edges, from the point cloud information collected by the LiDAR and register them with the point cloud model obtained based on the image.

[0063] The purpose of point cloud data fusion is: the point cloud information obtained by the lidar is accurate but cannot identify the type of object, and the image information obtained by the camera is rich but the positioning accuracy is not high. Therefore, this technical feature fuses the data from different sensors to obtain a more accurate and effective three-dimensional point cloud model.

[0064] In the process of constructing the 3D mesh, the fused point cloud is further processed, such as removing duplicate points, filling holes, etc., connecting the point cloud data into a triangular mesh, and converting the discrete point cloud data into a continuous surface model;

[0065] In the process of building a model, you need to go through the following steps:

[0066] (1) Rigid body transformation: that is, the transformation from the world coordinate system to the camera coordinate system. This process changes the spatial position and direction of the object, but keeps the shape and size of the object unchanged. The coordinates of the object in the world coordinate system (x w ,y w , z w ) and the coordinates in the camera coordinate system (x c ,y c , z c ), then the relationship between them can be described by the rotation matrix R and the translation vector T, which satisfies the following formula:

[0067]

[0068] Where R is the 3×3 rotation matrix, T is the 3×1 translation vector, and 0 is the 1×3 zero vector.

[0069] (2) Perspective projection: This is the transformation from the camera coordinate system to the image coordinate system. It can also be understood as the process of mapping points in the three-dimensional camera coordinate system to the two-dimensional image coordinate system. c ,y c , z c ) and the coordinates of the corresponding object image in the image coordinate system (x p ,y p ), according to the perspective projection principle, the following formula is satisfied:

[0070]

[0071] Where f is the focal length of the camera.

[0072] (3) Calculate the distortion model: Due to the physical characteristics of the camera lens, distortion will occur during the imaging process, so a distortion model needs to be introduced. The image coordinates before distortion removal are (x, y), and the image coordinates after distortion removal are (x crt ,y crt ). The distance r from a point (x, y) in the image to the center of the image satisfies the following formula:

[0073]

[0074] Distortion includes radial distortion and tangential distortion. The coordinates after distortion elimination (x crt ,y crt ) satisfies the following formula:

[0075]

[0076] Among them, x rcrt and rcrt is the radial distortion term, satisfying the following formula:

[0077]

[0078] x tcrt and tcrt is the tangential distortion term, satisfying the following formula:

[0079]

[0080] Among them, k1, k2, k3, p1, and p2 are all distortion parameters.

[0081] (4) Secondary transformation: that is, the transformation from the image coordinate system to the pixel coordinate system. The coordinates of the object in the image coordinate system (x p ,y p ) and the coordinates (u, v) of the corresponding pixel in the digital image satisfy the following formula:

[0082]

[0083] Among them, (u0, v0) is the coordinate of the image center in the pixel coordinate system.

[0084] In summary, the imaging model of the camera is established, and the transformation from the pixel coordinate system to the world coordinate system satisfies the following formula:

[0085]

[0086] in, is the camera’s intrinsic parameter matrix, is the extrinsic parameter matrix of the camera.

[0087] In order to obtain the camera's intrinsic matrix, extrinsic matrix, and distortion parameters, it is necessary to use the camera to extract information from the calibration object. By finding a large number of points (x w ,y w , z w ) and its corresponding point (u, v), solve the unknown parameters in the model, and obtain an accurate imaging model through camera calibration, which can map each pixel in the image back to the corresponding 3D world coordinates.

[0088] The 3D grid model is projected onto a 2D plane to form a 2D local road map, which identifies information such as movable areas, obstacles and other road participants. While building the map, the SLAM algorithm needs to achieve real-time positioning, which is achieved through map matching and motion model estimation. Map matching is to match the current sensor data with the constructed map to determine the location of the device in the map.

[0089] Motion model estimation is to combine the data of the sensor system and use the motion model to estimate the motion trajectory and posture of the device. According to the motion equation and observation equation, the vector representing the position coordinate of the robot is denoted as x, which can be used in the three-dimensional coordinate system as [x1,x2,x3] T To represent, let the position of the landmark be y j , where j is the landmark number, and the robot's internal motion sensor reading is recorded as u, which is the amount of motion. For example, it can be measured by displacement sensors and IMU [Δx1, Δx2, Δθ] T As the motion amount u at a certain moment, the measurement amount of the observed landmark obtained by the camera, lidar, etc. is recorded as z j , where j is the landmark number and k is the time series number. The formula is as follows:

[0090]

[0091] The error can be expressed as:

[0092]

[0093] The second norm of minimizing the error is expressed as:

[0094]

[0095] In the process of local path planning, based on the constructed three-dimensional map model, the path planning system dynamically plans the specific operation path of the two-wheeled electric vehicle. The specific local path planning process is as follows:

[0096] 1. Motion state solution: Based on the three-dimensional environment model obtained in step 2, combined with the data of the speed sensor and inertial sensor, the two-dimensional motion trajectory of the two-wheeled electric vehicle, the two-wheeled electric vehicle's two-dimensional coordinates, the speed and acceleration of the two-wheeled electric vehicle and other motion information are solved.

[0097] 2. Frenet coordinate conversion: Based on the route obtained in step 1, set the center line of the road and construct the Frenet coordinate system. The coordinates of the two-wheeled electric vehicle in the Frenet coordinate system can be solved according to the following equations:

[0098]

[0099]

[0100] x=x r -d sinΔθ r

[0101] y=y r +d cos Δθ r

[0102]

[0103]

[0104] 3. Lateral trajectory planning: During the automatic driving of a two-wheeled electric vehicle, the lateral movement relative to the center line of the road is mainly responsible for obstacle avoidance and turning. The lateral displacement state of the two-wheeled electric vehicle at the current moment is known. The present invention adopts a fifth-order polynomial as a candidate function of the lateral movement amount:

[0105] d(t)=c d0 +c d1 t+c d2 t 2 +c d3 t 3 +c d4 t 4 +c d5 t 5

[0106]

[0107]

[0108] 4. Longitudinal trajectory planning: During the automatic driving of a two-wheeled electric vehicle, the longitudinal movement along the center line of the road is mainly responsible for following the vehicle and parking. The longitudinal displacement state of the two-wheeled electric vehicle at the current moment is known. The present invention adopts a quartic polynomial as a candidate function of the longitudinal movement amount:

[0109] s(t)=c s0 +c s1 t+c s2 t 2 +c s3 t 3 +c s4 t 4

[0110]

[0111]

[0112] 5. Collision detection: Generate candidate trajectories based on the above polynomials. If the two-wheeled electric vehicle travels a certain distance along the candidate trajectory toward the center line of the road within a sampling time, and the distance between the two-wheeled electric vehicle and any obstacle is less than the set threshold after the end, the collision detection module will eliminate the candidate curve.

[0113] 6. Solving the optimal trajectory: Let the lateral / longitudinal movement of the two-wheeled electric vehicle in the Frenet coordinate system be f(t), and construct a one-dimensional integral system for the trajectory of the two-wheeled electric vehicle:

[0114]

[0115] in One of the optimization goals of path planning is to reduce the change in the rate of change of acceleration f(t), that is, to solve the above integral system so that Minimum.

[0116] In the process of steering control, according to the determined path and the current state of the vehicle, the control system controls the front wheel steering angle to ensure that the two-wheeled electric vehicle moves along the path planned in step 3. The specific steps of steering control are as follows:

[0117] 1. Calculate the heading angle: Calculate the tangent direction of the two-wheeled electric vehicle path trajectory obtained in step 3 at the current path point, and make an angle with the movement direction of the two-wheeled electric vehicle returned by the sensor to obtain the heading angle θ.

[0118] 2. Calculation of electric vehicle steering angle: The two-wheeled electric vehicle can be simplified as a rigid crossbar with a length of L in terms of force motion. The force point is the connection between the rear wheels. Assuming the current steering angle is δ, the rotation radius of the two-wheeled electric vehicle is R = L / tan(δ); according to the instantaneous rotation law, if the real-time speed is v, the rotation rate is The rate of change of the rotation speed can be obtained by the second-order differential of the path trajectory, and then the steering angle δ and its rate of change can be solved The torque motor is then used to apply torque to the front wheels at a rate of change. Change the direction of the front wheels.

[0119] 3. Adjust the steering speed: the maximum speed at which the electric vehicle will not slip when turning, taking into account the maximum static friction force F between the tire and the ground f =μ s mg, where μ s is the friction coefficient between the tire and the ground. Considering that the friction coefficient of a general road is 0.4-0.6, the invention stipulates that the maximum centrifugal force during steering shall not be greater than 0.3 mg. The maximum driving speed can be calculated as If the electric vehicle speed is higher than the maximum turning speed, the vehicle speed is reduced by reducing the electric motor power and slowly activating the brakes.

[0120] The specific steps of electric vehicle self-balancing driving are:

[0121] 1. Static balance: When the electric vehicle is in standby mode, the auxiliary wheels deployed on both sides work together with the stable tripod structure to achieve and maintain the stability and balance of the vehicle body. This design ensures that the electric vehicle can be stably parked when not moving to prevent accidental tipping or sliding.

[0122] 2. Electric vehicle start: After the autonomous driving path is set and confirmed, the electric vehicle immediately executes the start procedure. During this process, the tripod is automatically retracted to prepare for driving, and then the electric vehicle starts the motor and gradually increases to the preset minimum driving speed. Once the speed threshold is reached, in order to enhance driving stability and efficiency, the electric vehicle will further retract the auxiliary wheels on both sides, thus entering the normal driving state.

[0123] 3. Driving speed control: According to the current motion state of the electric vehicle, calculate the current minimum / maximum driving speed, and control the electric vehicle according to the following logic:

[0124] ① If the speed of the electric vehicle is lower than the minimum driving speed, the electric motor is controlled to provide maximum power to the electric vehicle;

[0125] ② If the speed of the electric vehicle is greater than the minimum driving speed and lower than the maximum driving speed, the power required to maintain the maximum driving speed is provided;

[0126] ③ If the speed of the electric vehicle is greater than the maximum driving speed, the brakes are controlled to slow down the electric vehicle.

[0127] 4. Electric vehicle balancing system: Considering that the ratio of the load to the deadweight of a two-wheeled electric vehicle is much higher than that of a car, and its center of gravity also changes with the change of the driver's posture, it is difficult to ensure the balance of the electric vehicle by controlling the driving speed alone. Therefore, the present invention adds a self-balancing system based on a gyroscope and a flywheel to the two-wheeled electric vehicle:

[0128] ① High-precision gyroscopes and accelerometers are installed in the center of the electric vehicle body to measure the dynamic parameters of the electric vehicle such as lateral tilt angle, angular velocity and acceleration in real time and accurately. The sensor data is transmitted to the control system via wired mode to ensure the real-time and reliability of data transmission.

[0129] ②The STM32-based computing platform is responsible for receiving data from the attitude sensor and quickly processing and analyzing it. The computing platform calculates the correction torque size and direction required by the current electric vehicle and the action instructions for the deflection motor according to the preset control algorithm.

[0130] ③The yaw motor is installed at an appropriate position of the electric vehicle and connected to a flywheel that can rotate at high speed through a mechanical transmission device. The flywheel is designed to have a high moment of inertia to store or release energy and generate a stable correction torque. The control system controls the rotation speed and direction of the flywheel by adjusting the current, voltage and frequency of the yaw motor, thereby achieving rapid adjustment of the tilt state of the electric vehicle.

[0131] 5. When an electric vehicle is driving and encounters a situation where it needs to stop, such as when it encounters a red light or reaches a preset destination, the vehicle will shut down the electric motor, then deploy the auxiliary wheels on both sides to enhance stability, and simultaneously control the brake system to gradually slow down the speed until the electric vehicle stops completely, and then deploy the supporting legs to keep the electric vehicle stationary.

[0132] In global route planning, a road graph model is constructed based on satellite images, city maps and other data. In the graph model, buildings, roads, public facilities, etc. in the city are regarded as nodes, and each node has a unique identifier and corresponding attributes.

[0133] Identifiers and corresponding attributes include name, position, height and type, and define the relationships between nodes, such as adjacent relationships, connection relationships, etc. These relationships can be represented in the graph model in the form of edges.

[0134] The navigation system uses the wireless communication module to request the navigation map from the cloud server, and obtains the basic information of the electric vehicle's position and direction in the map in real time through the GPS module.

[0135] Various traffic signs and signal light status can be identified through the camera.

[0136] The local path planning has the following limitations: the driving path should be limited to the road range planned in step 1; the next path point should be within a fan of a given angle centered on the current vehicle head direction; any path point cannot be within a safe distance from a stationary obstacle; when reaching any path point, it cannot overlap with the predicted path of a moving obstacle in the current environment; the moving path needs to meet the smoothness requirement, and the product of the curvature and the speed cannot be greater than the limit value to ensure the balance of the two-wheeled electric vehicle.

[0137] When calculating the route in global route planning, 3-5 candidate routes with the shortest time and closest distance are calculated. Road information such as road flow and traffic speed are collected based on the Internet of Vehicles system, and real-time weather information is obtained from local meteorological stations. The above information is analyzed to select the best route.

[0138] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the technology may modify or change the above embodiments without violating the spirit and scope of the present invention. Therefore, all equivalent modifications or changes made by those with ordinary knowledge in the technical field without departing from the spirit and technical ideas disclosed by the present invention should still be covered by the claims of the present invention.

Claims

1. An automatic driving system for a two-wheeled electric vehicle, characterized in that: Including navigation system, sensor system, computing platform, power system, and self-balancing system; The navigation system is composed of a GPS module and a wireless communication module; The sensor system is composed of a laser radar, a camera, an acceleration sensor and a gyroscope sensor; The computing platform is composed of a target recognition module, a three-dimensional reconstruction module, a path planning module and a control module; The power system is composed of a torque motor, an electric motor and a brake; The self-balancing system consists of a deflection motor, a flywheel, a supporting tripod and auxiliary wheels.

2. The automatic driving system for a two-wheeled electric vehicle according to claim 1, further comprising an automatic driving method, characterized in that: The automatic driving method comprises the following steps: Step 1: Global route planning The global route planning is performed by locating the position of the vehicle itself with GPS, and using the wireless communication module to request navigation services from the cloud server. The cloud server plans the route of the two-wheeled electric vehicle in the city map based on satellite positioning, and calculates the shortest and closest route based on the road traffic, speed, environment and weather conditions of the Internet of Vehicles system, and determines and returns the driving range and driving direction of the two-wheeled electric vehicle. Step 2: Real-time collection of road information The laser radar obtains real-time information about obstacles on the road, and the camera obtains real-time visual information of the road scene. At the same time, the acceleration sensor and gyroscope sensor record the posture changes of the electric vehicle, including acceleration and angular velocity, to provide auxiliary information for positioning; After acquiring and processing the sensor data, the SLAM algorithm begins to build a road environment model to plan the electric vehicle's driving path in the future, and builds a three-dimensional network model. The three-dimensional grid model is projected onto a two-dimensional plane to form a two-dimensional local road map, which identifies the movable area, obstacles and other road participant information. The SLAM algorithm realizes real-time positioning of the electric vehicle, matches the sensor data with the constructed map, and determines the location of the device in the map; Step 3: Local path planning Based on the constructed three-dimensional map model, the path planning system dynamically plans the specific operation path of the two-wheeled electric vehicle; Step 4: Steering Control According to the determined path and the current state of the vehicle including speed and heading angle, the control system of the present invention ensures that the two-wheeled electric vehicle moves along the planned path by controlling the front wheel steering angle; Step 5: Self-balancing electric vehicle The deflection motor, flywheel, support tripod and auxiliary wheels cooperate with the gyroscope to ensure the stability of the electric vehicle during turning, driving, driving at the lowest speed, driving at the highest speed and waiting for parking during the automatic driving process.

3. The automatic driving method of a two-wheeled electric vehicle according to claim 2, characterized in that: In global route planning, a road graph model is constructed based on satellite images, city maps and other data. In the graph model, buildings, roads, public facilities, etc. in the city are regarded as nodes, and each node has a unique identifier and corresponding attributes.

4. The automatic driving method for a two-wheeled electric vehicle according to claim 3, characterized in that: Identifiers and corresponding attributes include name, position, height and type, and define the relationships between nodes, such as adjacent relationships, connection relationships, etc. These relationships can be represented in the graph model in the form of edges.

5. The automatic driving method of a two-wheeled electric vehicle according to claim 2, characterized in that: The navigation system uses the wireless communication module to request the navigation map from the cloud server, and obtains the basic information of the electric vehicle's position and direction in the map in real time through the GPS module.

6. The automatic driving method of a two-wheeled electric vehicle according to claim 2, characterized in that: Various traffic signs and signal light status can be identified through the camera.

7. The automatic driving method for a two-wheeled electric vehicle according to claim 2, characterized in that: The local path planning has the following limitations: the driving path should be limited to the road range planned in step 1; the next path point should be within a fan of a given angle centered on the current vehicle head direction; any path point cannot be within a safe distance from a stationary obstacle; when reaching any path point, it cannot overlap with the predicted path of a moving obstacle in the current environment; the moving path needs to meet the smoothness requirement, and the product of the curvature and the speed cannot be greater than the limit value to ensure the balance of the two-wheeled electric vehicle.

8. The automatic driving method of a two-wheeled electric vehicle according to claim 2, characterized in that: When calculating the route in global route planning, 3-5 candidate routes with the shortest time and closest distance are calculated. Road information such as road flow and traffic speed are collected based on the Internet of Vehicles system, and real-time weather information is obtained from local meteorological stations. The above information is analyzed to select the best route.