Driverless car coupling control system based on kinetic model

Abstract

The invention provides a driverless automobile coupling control system based on a kinetic model, and belongs to the technical field of new energy automobiles, and the driverless automobile coupling control system comprises a sensor module which is used for obtaining the driving state of a driverless automobile; the bicycle kinetic model modeling module is used for establishing a kinetic model of the pilotless automobile; the steering control module is used for calculating a tire slip angle during steering based on a tire rigidity model and adjusting a steering torque; the coupling control module is used for generating control instructions for steering, acceleration and speed of the pilotless automobile based on the bicycle model created by the bicycle dynamics model modeling module and the tire rigidity model used by the steering control module; the control strategy is completed by combining a linear quadratic regulator and a model prediction controller in the advanced control field, accurate control over the pilotless automobile is achieved, static and dynamic obstacles can be predicted and avoided in real time, and therefore the driving stability and safety of the automobile are remarkably improved.

Description

Technical Field

[0001] The present invention mainly relates to the technical field of new energy vehicles, and in particular to a driverless vehicle coupling control system based on a dynamic model. Background Art

[0002] With the rapid development of autonomous driving technology, the autonomous control system of vehicles has become an important part of the intelligent transportation system. However, traditional vehicle control systems usually rely only on simple dynamic models and a single control strategy, making it difficult to achieve high-precision trajectory tracking and safe obstacle avoidance in complex and changing road environments. Existing systems often face challenges such as slow response speed and insufficient control accuracy when dealing with issues such as vehicle lateral stability, longitudinal control, and dynamic obstacle avoidance, making it difficult to fully ensure driving safety.

[0003] In complex road conditions, the dynamic behavior of the vehicle is closely related to factors such as tire stiffness and vehicle steering. Traditional lateral control methods fail to fully consider the impact of tire stiffness on steering, which makes the vehicle prone to instability at high speed or sharp turns. In addition, longitudinal control usually uses a single-loop PID controller, which can only simply adjust the speed and cannot accurately respond to distance control requirements, especially in vehicle following scenarios, where the response to changes in distance and speed of the vehicle in front is delayed. Current obstacle recognition technologies mostly rely on a single sensor and fail to fully utilize the advantages of multiple sensor fusion, limiting the safety and intelligent decision-making capabilities of vehicles in complex environments. In response to this challenge, some unmanned driving manufacturers have systematically tested the trajectories and obstacle avoidance of multiple unmanned vehicles. However, how to efficiently achieve obstacle avoidance of static and dynamic obstacles in actual application scenarios is still a great challenge, and the safety risks in the test will also affect the efficiency of unmanned vehicles when driving.

[0004] Therefore, it is urgent to design a safety-oriented unmanned vehicle coupling control system based on a dynamic model to solve the above-mentioned technical problems. Summary of the invention

[0005] The present invention mainly provides an unmanned vehicle coupling control system based on a dynamic model to solve the technical problems raised in the above background technology.

[0006] The technical solution adopted by the present invention to solve the above technical problems is:

[0007] A coupling control system for an unmanned vehicle based on a dynamic model, comprising:

[0008] The sensor module is used to obtain the driving status of the driverless car, including the parameters such as vehicle speed, steering angle, and lateral acceleration;

[0009] The bicycle dynamics model modeling module builds the dynamics model of the driverless car based on the parameters collected by the sensor module. The model simplifies the driverless car into a bicycle model with the front wheel controlling the steering and the rear wheel controlling the drive.

[0010] The steering control module calculates the tire slip angle during steering based on the tire stiffness model and adjusts the steering torque to optimize the vehicle's turning stability and safety;

[0011] The coupling control module generates control instructions for the steering, acceleration and speed of the driverless car based on the bicycle model created by the bicycle dynamics model modeling module and the tire stiffness model used by the steering control module, thereby improving the vehicle's handling performance and safety.

[0012] Furthermore, in the present invention, the system implements lateral control of the driverless car through a linear quadratic regulator controller, and the lateral control includes:

[0013] Obtain the lateral offset and heading angle error of the vehicle based on the bicycle model;

[0014] The linear quadratic regulator controller calculates the optimal steering angle in real time according to the offset and the error to achieve lateral stability and trajectory tracking of the driverless car.

[0015] Furthermore, in the present invention, the system also includes a dual-loop PID controller for longitudinal control of the driverless vehicle, and the longitudinal control includes:

[0016] The first PID control loop is used to control the distance between the vehicle and the target in front, and to maintain a safe distance by adjusting the throttle and brake;

[0017] The second PID control loop is used to perform speed control based on the deviation between the current speed of the vehicle and the preset speed, and to achieve precise speed control by adjusting the throttle or brake.

[0018] Furthermore, in the present invention, the dual-loop PID controller adjusts the acceleration and deceleration of the vehicle in real time to ensure the longitudinal control stability of the vehicle under complex road conditions, and dynamically adjusts the speed and distance control parameters according to traffic conditions.

[0019] Furthermore, in the present invention, the system also implements tracking control of the driverless car through a model predictive controller, and the tracking control includes:

[0020] Predict the vehicle trajectory at future moments based on the vehicle’s current state, target trajectory, and dynamics model;

[0021] The model predictive controller calculates the optimal steering, acceleration, and braking operations by optimizing the control inputs to keep the vehicle along the target trajectory while satisfying the vehicle dynamics constraints and safety constraints.

[0022] Furthermore, in the present invention, the model predictive controller can update the target trajectory according to real-time environmental changes and vehicle status, and adjust the control input to ensure vehicle tracking accuracy and safety while taking into account road curvature, obstacles and vehicle dynamics.

[0023] Furthermore, in the present invention, the system also includes an obstacle recognition module, which realizes obstacle detection through a sensor installed on the top of the driverless car, and the obstacle detection includes:

[0024] The sensors detect the environment within 50 meters around the driverless car;

[0025] Up to 20 targets can be identified and corresponding obstacle information can be generated based on the distance, speed and direction of the target.

[0026] Furthermore, in the present invention, the sensor includes a laser radar, a millimeter wave radar and a camera to improve the accuracy of obstacle recognition and environmental adaptability, especially to achieve accurate detection of dynamic and static obstacles under complex road conditions.

[0027] Furthermore, in the present invention, the system can achieve static obstacle avoidance by generating an ST graph (space-time graph), wherein the static obstacle avoidance includes:

[0028] A dynamic programming algorithm is used to plan the optimal driving path based on the obstacle positions and path information in the ST diagram to ensure that the vehicle bypasses static obstacles.

[0029] Furthermore, in the present invention, the system can avoid dynamic obstacles through a quadratic programming algorithm, and the dynamic obstacle avoidance includes:

[0030] Combined with the time dimension in the ST diagram, the moving trajectory and speed of the dynamic obstacle are calculated, and a quadratic programming model is generated to adjust the vehicle's trajectory in real time to ensure that the vehicle's driving stability and safety are maintained while avoiding dynamic obstacles.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] The present invention uses a coupling control system based on a dynamic model, combined with a linear quadratic regulator and a model predictive controller in the field of advanced control to complete the control strategy, as well as dynamic programming and quadratic programming algorithms, to achieve precise control and safe driving of driverless cars, and can predict and avoid static and dynamic obstacles in real time, thereby significantly improving the driving stability and safety of the vehicle. In addition, the system's multi-sensor fusion and dynamic programming algorithms effectively enhance environmental perception capabilities, ensuring that the vehicle can complete tasks safely and smoothly in complex traffic environments, and improving the practicality and reliability of autonomous driving technology.

[0033] The present invention will be explained in detail below with reference to the accompanying drawings and specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 The figure is a flow chart of the technical solution for a coupling control system of an unmanned vehicle based on a dynamic model. DETAILED DESCRIPTION

[0035] To facilitate understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in different forms and is not limited to the embodiments described in the text. On the contrary, these embodiments are provided to make the content disclosed in the present invention more thorough and comprehensive.

[0036] It should be noted that when an element is referred to as being "fixed on" another element, it may be directly on the other element or there may be a central element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be a central element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used in this article are for illustrative purposes only.

[0037] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which the present invention belongs. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0038] For examples, please refer to the attached Figure 1 , a driverless car coupling control system based on a dynamic model, comprising:

[0039] The sensor module is used to obtain the driving status of the driverless car, including the parameters such as vehicle speed, steering angle, and lateral acceleration;

[0040] The bicycle dynamics model modeling module builds the dynamics model of the driverless car based on the parameters collected by the sensor module. The model simplifies the driverless car into a bicycle model with the front wheel controlling the steering and the rear wheel controlling the drive.

[0041] The steering control module calculates the tire slip angle during steering based on the tire stiffness model and adjusts the steering torque to optimize the vehicle's turning stability and safety;

[0042] The coupling control module generates control instructions for the steering, acceleration and speed of the driverless car based on the bicycle model created by the bicycle dynamics model modeling module and the tire stiffness model used by the steering control module, thereby improving the vehicle's handling performance and safety.

[0043] It should be noted that, in this embodiment, the method used by the system includes the following steps:

[0044] S1: First, various sensors in the sensor module (such as GPS, IMU, speed sensor, angle sensor, etc.) are used to collect the vehicle's speed, steering angle, lateral acceleration, longitudinal acceleration and other parameters related to the vehicle's dynamic state. The data collected by different sensors are fused, and algorithms such as Kalman filtering are used to ensure that the acquired vehicle status information is accurate and reliable, and eliminate the influence of sensor noise, providing high-quality data support for subsequent control.

[0045] S2: Based on the parameters related to the vehicle dynamic state collected in step S1, a dynamic model of the driverless car is established. The model simplifies the driverless car into a bicycle model with the front wheel controlling the steering and the rear wheel controlling the drive. Then, a state equation is established. Through physical laws and geometric relationships, a state equation describing the vehicle motion is established, covering the lateral and longitudinal motion behaviors of the vehicle, including dynamic characteristics such as position, speed, and acceleration. Finally, the established dynamic model is verified and compared with the driving data of the actual vehicle to ensure the accuracy and applicability of the model, providing a basis for subsequent control strategies.

[0046] The modeling formula of the kinetic model is:

[0047]

[0048] Among them, x and y represent the lateral and mirror displacement of the driverless car in the horizontal coordinate system. represents the heading angle, β represents the sideslip angle, δ f ,δ r They represent the turning angles of the front and rear wheels of the driverless car respectively, v represents the speed of the driverless car, and L represents the wheelbase of the driverless car.

[0049] S3: Based on the dynamic model established in step S3, consider the influence of tire stiffness on the steering of the driverless car. First, establish a tire stiffness model and analyze the side slip characteristics under different tire conditions to evaluate the influence of the tire on the dynamic behavior of the vehicle during steering. Then calculate the side slip angle. During the steering process, calculate the side slip angle of the tire based on the current speed and steering angle, and then derive the demand for steering torque. Finally, perform control optimization and adjust the steering torque in real time to optimize the stability and safety of the vehicle when turning, ensure that the vehicle is not prone to loss of control at high speed or sharp turns, and can accurately follow the predetermined trajectory. The modeling formula of the dynamic model after considering tire stiffness is:

[0050]

[0051] Among them C αf ,C αr They represent the stiffness of the front and rear tires, I represents the moment of inertia of the tire, m represents the mass of the tire, a and b represent the front and rear wheelbases of the driverless car, and v x represents the radial velocity in the x direction, and δ is the front wheel turning angle of the driverless car.

[0052] S4: Based on steps S1 to S3, the bicycle model and the tire stiffness model are integrated to generate control instructions for the driverless car. The specific functions include:

[0053] Control command generation: Generate control commands for vehicle steering, acceleration and speed based on the status information output by the dynamic model to ensure that the vehicle can drive safely and smoothly under various driving conditions;

[0054] Real-time monitoring and adjustment: During actual driving, the deviation between the vehicle status and the preset target is monitored in real time, and control instructions are adjusted dynamically to respond to emergencies or environmental changes, thereby improving the vehicle's handling performance and safety.

[0055] S5: Based on steps S1 to S4, the lateral offset and heading angle error of the driverless car are obtained. The system uses a bicycle model to obtain the lateral offset and heading angle error of the vehicle in real time, which specifically includes the following aspects:

[0056] The sensor module collects the geometric relationship between the vehicle's current position and the planned driving path, and calculates the vehicle's lateral offset. This offset represents the degree of deviation of the vehicle's current position from the ideal trajectory and is a key input for optimizing the control strategy.

[0057] By comparing the actual heading angle of the vehicle with the expected heading angle, the heading angle error is calculated. This error reflects the deviation between the vehicle's orientation and the target path, which helps to determine whether the vehicle has deviated from the predetermined trajectory;

[0058] The calculation of lateral offset and heading angle error adopts a real-time update mechanism to ensure that dynamic changes can be quickly reflected during vehicle driving. This mechanism achieves high-frequency status monitoring through fast sampling and data processing, providing basic data support for subsequent control decisions.

[0059] S6: Based on step S5, after obtaining the lateral offset and heading angle error, the LQR controller calculates the optimal steering angle in real time based on this information. The specific steps include:

[0060] Based on the bicycle model, a state space model describing the dynamic behavior of the vehicle is established. The model takes the lateral offset and heading angle error as input variables and the steering angle as output variable to form a complete control system framework.

[0061] The LQR controller weighs the impact of lateral offset and heading error on vehicle control by defining a cost function. The cost function is usually composed of a weighted combination of control energy and deviation degree, ensuring that the intensity of control input is reduced while pursuing trajectory tracking accuracy. The cost function is expressed as follows:

[0062]

[0063] Among them, J is the cost function itself, e rr is the error function, u is the loss function, Q and R are the parameter matrices of the linear quadratic regulator (LQR);

[0064] By solving the Riccati equation, the LQR controller calculates the state feedback gain matrix and adjusts the control strategy in real time to achieve the optimal steering angle input. The gain matrix is ​​used to adjust the vehicle's steering behavior to ensure that the vehicle can drive towards the predetermined target trajectory at any moment.

[0065] In real-time applications, the LQR controller not only calculates the steering angle, but also evaluates the vehicle's dynamic response to ensure that the vehicle maintains lateral stability and maneuverability under rapidly changing conditions. Through a continuous feedback mechanism, the LQR controller can dynamically adapt to various driving environments, improve the vehicle's trajectory tracking capabilities, and ultimately achieve a safe and smooth autonomous driving experience.

[0066] S7: This step aims to achieve distance control based on the vehicle and the target ahead through the first PID control loop to ensure a safe vehicle distance, including:

[0067] Use front sensors (such as lidar or ultrasonic sensors) to measure the distance between the vehicle and the front target (such as the vehicle in front) in real time to achieve distance control, ensuring that the latest distance information can be obtained quickly and accurately;

[0068] The distance error is calculated based on the difference between the currently measured distance and the preset safety distance. This error will be used as the input of the PID controller to indicate the control amount that needs to be adjusted;

[0069] Design the first PID controller and set the proportional, integral and differential parameters to respond to the distance error. The proportional term is used to adjust the throttle or brake in real time, the integral term eliminates long-term static errors, and the differential term is used to predict the trend of distance changes and enhance the responsiveness of the system;

[0070] According to the control signal output by the PID controller, the throttle and brake are adjusted in real time to ensure that a safe distance is always maintained between the vehicle and the vehicle in front. When the distance to the vehicle in front is too close, the control system will automatically reduce the throttle or apply the brakes, otherwise it will increase the throttle to maintain smooth driving.

[0071] S8: Based on step S7, the second PID control loop is used to perform speed control based on the deviation between the current speed of the vehicle and the preset speed. The specific operations include:

[0072] The current speed is obtained in real time through the vehicle's internal sensors (such as vehicle speed sensors) and compared with the set target speed to ensure the accuracy of speed feedback;

[0073] Calculating an error between a current speed and a preset target speed as an input to a second PID controller, the error indicating a deviation between a current driving state of the vehicle and a desired driving state;

[0074] Design the second PID controller and set appropriate proportional, integral and differential parameters to ensure flexibility and accuracy of speed control. The proportional term will directly affect the adjustment of the current speed, the integral term is used to correct long-term deviations, and the differential term helps to reduce the suddenness of speed changes;

[0075] According to the output signal of the second PID controller, the throttle and brake are adjusted in real time to ensure that the vehicle travels at the set speed. When the speed is insufficient, the system will increase the throttle to increase the speed, and when the speed is over, the brake will be applied in time to ensure that the vehicle always runs within a safe and stable speed range.

[0076] S9: Based on steps S1 to S8, in this step, the system predicts the vehicle trajectory at a future time according to the vehicle's current state, target trajectory, and dynamic model, specifically including:

[0077] The system collects the vehicle's current status information in real time, including position, speed, steering angle, and acceleration, to ensure a comprehensive understanding of the vehicle's dynamic behavior;

[0078] Setting the target trajectory, usually generated by a predetermined navigation path or driving instructions, ensures that the vehicle moves in the desired direction;

[0079] The established dynamic model is used to simulate the movement of the vehicle at different time steps. By combining the current state with the target trajectory, the system can predict the vehicle's movement trajectory at future moments, including changes in position and speed. This prediction process takes into account the vehicle's dynamic characteristics, such as acceleration, steering ability, and the influence of the external environment, to ensure the accuracy and reliability of the model.

[0080] During the vehicle's driving, the system will continuously update the target trajectory and current status, and predict the future trajectory in real time to adapt to the dynamically changing driving environment. Based on the prediction results, the control system can make more accurate control decisions and improve the vehicle's driving safety and stability.

[0081] S10: Based on step S9, the control input is optimized by using a model predictive controller (MPC) to calculate the optimal steering, acceleration and braking operations to ensure that the vehicle travels along the target trajectory. The specific operations include:

[0082] MPC uses the predicted future trajectory and the vehicle's dynamic model to set the optimization problem of control input. The goal is to minimize the tracking error while satisfying the vehicle's dynamic constraints (such as maximum acceleration, steering angle limit, etc.) and safety constraints (such as safe distance from obstacles ahead);

[0083] Handling constraints: During the optimization process, MPC considers the physical limitations of the vehicle, such as maximum steering angle and acceleration capabilities, as well as the influence of the external environment, such as road curvature and obstacle locations. These constraints ensure that the vehicle is controlled within a safe range to avoid the risk of loss of control or collision.

[0084] By solving the optimization problem, MPC calculates the optimal steering angle, acceleration and braking commands in each control cycle. These commands will guide the vehicle towards the predicted trajectory, ensuring that the optimal driving path is maintained in a dynamic environment. Its cost function is expressed as:

[0085] J=w1J1+w2J2+w3J3

[0086] Among them, J is the cost function, J1, J2, J3 represent the cost similar to the original path, the smooth curve cost and the compact cost respectively, w1, w2, w3 are their respective weights;

[0087] During each control cycle, MPC will update the optimization strategy in real time based on the latest vehicle status information and external environment changes to ensure that the vehicle can always accurately track the target trajectory under changing conditions, thereby improving the responsiveness and safety of the unmanned driving system.

[0088] S11: In this step, the system detects the surrounding environment through the sensors installed on the top of the driverless car. The specific operations include:

[0089] The sensors are configured to scan a 50-meter range around the driverless car in real time to ensure that changes in the surrounding environment can be fully monitored;

[0090] Using a combination of sensors, such as lidar, cameras, and ultrasonic sensors, to provide information in different dimensions, including distance, speed, and object recognition, the system can improve the accuracy and reliability of environmental perception through this multi-sensor fusion approach;

[0091] The environmental data collected in real time is processed and analyzed to form a dynamic model of the surrounding environment for subsequent obstacle identification and path planning.

[0092] S12: In this step, obstacle information will be generated. The system can identify up to 20 targets through sensors installed on the top of the driverless car and generate corresponding obstacle information, including:

[0093] By analyzing the sensor data, the system can identify and classify surrounding obstacles, such as pedestrians, other vehicles, static obstacles, etc.

[0094] For each identified target, the system records key attributes such as distance, relative speed and orientation. This information is used to assess the impact of obstacles on the vehicle's driving path and perform subsequent dynamic planning;

[0095] All identified target information is integrated into an obstacle database, ensuring that subsequent algorithms can quickly access and utilize this data for real-time decision making.

[0096] S13: Based on steps S1 to S12, in this step, the system uses a dynamic programming algorithm to plan the optimal driving path according to the obstacle position and path information in the ST map. The specific operations include:

[0097] Constructing ST graph: Based on the obstacle positions detected by the sensors and the current state of the vehicle, a state-time (ST) graph is constructed, where nodes represent possible vehicle positions and edges represent feasible paths between these positions.

[0098] Using a dynamic programming algorithm, the optimal path from the starting position to the target position is found in the ST graph. The algorithm takes into account the impact of obstacles to ensure that the vehicle can effectively bypass static obstacles.

[0099] During the planning process, the system will evaluate the feasibility and safety of different paths in real time, and dynamically adjust the optimization results according to environmental changes (such as new obstacles) to ensure that the planned path is not only efficient, but also safe and feasible;

[0100] If new obstacles or environmental changes are detected during driving, the system will automatically trigger the re-planning mechanism and recalculate the optimal path to ensure that the vehicle can drive safely and stably in complex environments.

[0101] S14: Static obstacle avoidance is achieved according to step S13. In this step, the system predicts and adjusts the motion characteristics of dynamic obstacles in combination with the time dimension in the ST diagram. The specific operations include:

[0102] Dynamic obstacle recognition: First, the system uses sensors to identify dynamic obstacles around it, obtains key attributes such as its current position, speed, and direction of movement, and provides basic data for subsequent trajectory calculation;

[0103] Movement trajectory prediction: Based on the current speed and direction, the system uses motion models to predict the position changes of dynamic obstacles in the future. By analyzing the movement pattern of obstacles, the system can estimate their possible travel paths;

[0104] Quadratic programming model generation: Combined with the prediction information of dynamic obstacles, a quadratic programming model is generated. This model aims to optimize the vehicle's motion trajectory to ensure that the vehicle's driving stability and safety are maintained while avoiding dynamic obstacles;

[0105] Optimized control strategy calculation: In the quadratic programming model, the system uses mathematical optimization methods to calculate the optimal control inputs for the vehicle, including steering angle, acceleration and braking force. These control inputs will guide the vehicle to adjust its driving trajectory to avoid the predicted dynamic obstacles;

[0106] Real-time adjustment and feedback mechanism: During the driving process, the system continuously updates the position and speed information of obstacles and adjusts the optimization model in real time. If it detects that the behavior of dynamic obstacles has changed, the system will immediately recalculate the optimal trajectory to ensure that the vehicle can effectively avoid them while maintaining smooth driving;

[0107] Safety and stability assessment: During the path adjustment process, the system will continuously monitor the vehicle's dynamic performance, assess the vehicle's stability and safety, and ensure that the adjusted trajectory will not cause the vehicle to lose control or cause other safety hazards.

[0108] Furthermore, the system implements lateral control of the driverless car through a linear quadratic regulator controller, and the lateral control includes:

[0109] Obtain the lateral offset and heading angle error of the vehicle based on the bicycle model;

[0110] The linear quadratic regulator controller calculates the optimal steering angle in real time according to the offset and the error to achieve lateral stability and trajectory tracking of the driverless car.

[0111] Furthermore, the system also includes a dual-loop PID controller for longitudinal control of the driverless vehicle, wherein the longitudinal control includes:

[0112] The first PID control loop is used to control the distance between the vehicle and the target in front, and to maintain a safe distance by adjusting the throttle and brake;

[0113] The second PID control loop is used to perform speed control based on the deviation between the current speed of the vehicle and the preset speed, and to achieve precise speed control by adjusting the throttle or brake.

[0114] Furthermore, the dual-loop PID controller adjusts the acceleration and deceleration of the vehicle in real time to ensure the longitudinal control stability of the vehicle under complex road conditions, and dynamically adjusts the speed and distance control parameters according to traffic conditions.

[0115] Furthermore, the system also implements tracking control of the driverless car through a model predictive controller, and the tracking control includes:

[0116] Predict the vehicle trajectory at future moments based on the vehicle’s current state, target trajectory, and dynamics model;

[0117] The model predictive controller calculates the optimal steering, acceleration, and braking operations by optimizing the control inputs to keep the vehicle along the target trajectory while satisfying the vehicle dynamics constraints and safety constraints.

[0118] Furthermore, the model predictive controller can update the target trajectory according to real-time environmental changes and vehicle status, and adjust the control input to ensure vehicle tracking accuracy and safety under the premise of considering road curvature, obstacles and vehicle dynamics;

[0119] It should be noted that, in this embodiment, compared with the traditional unmanned driving control module, the coupling control module of the present invention combines the LQR and MPC control strategies, and the coupling control module can better balance the lateral and longitudinal control, provide more precise control capabilities, and adapt to various driving scenarios; at the same time, the module has real-time monitoring and feedback capabilities, and can quickly adjust the control strategy according to the vehicle status and environmental changes, thereby improving the vehicle's responsiveness and stability in complex environments; secondly, the module can flexibly adjust the control strategy under different traffic conditions and environmental conditions, and is suitable for a variety of scenarios, such as cities, rural areas or highways.

[0120] Furthermore, the system also includes an obstacle recognition module, which realizes obstacle detection through a sensor installed on the top of the driverless car, and the obstacle detection includes:

[0121] The sensors detect the environment within 50 meters around the driverless car;

[0122] Up to 20 targets can be identified and corresponding obstacle information can be generated based on the distance, speed and direction of the target.

[0123] Furthermore, the sensor includes a laser radar, a millimeter-wave radar and a camera to improve the accuracy of obstacle recognition and environmental adaptability, especially to achieve accurate detection of dynamic and static obstacles under complex road conditions.

[0124] Furthermore, the system can avoid static obstacles by generating an ST graph (space-time graph), wherein the static obstacle avoidance includes:

[0125] A dynamic programming algorithm is used to plan the optimal driving path based on the obstacle positions and path information in the ST diagram to ensure that the vehicle bypasses static obstacles.

[0126] Furthermore, the system can avoid dynamic obstacles through a quadratic programming algorithm, and the dynamic obstacle avoidance includes:

[0127] Combined with the time dimension in the ST diagram, the moving trajectory and speed of the dynamic obstacle are calculated, and a quadratic programming model is generated to adjust the vehicle's trajectory in real time to ensure that the vehicle's driving stability and safety are maintained while avoiding dynamic obstacles.

[0128] For details, please refer to the attached Figure 1 , simplify the specific steps into four steps:

[0129] Step 1: Collect parameters such as vehicle speed, steering angle, lateral acceleration, etc. to monitor the vehicle dynamic state in real time, which are used to calculate the parameters of the unmanned driving controller;

[0130] Step 2: Based on the parameters collected by the sensor module in Step 1, a dynamic model of the driverless car is established. This model simplifies the vehicle into a form where the front wheels control steering and the rear wheels control driving, and takes into account factors such as the center of gravity, front and rear wheelbase, and vehicle mass of the vehicle to improve the accuracy of the simulation;

[0131] Step 3: According to the parameters collected by the sensor module in Step 1, the steering control module taking into account the tire stiffness calculates the tire slip angle during steering based on the tire stiffness model, and adjusts the steering torque in real time to optimize the turning stability of the vehicle under different driving conditions;

[0132] Step 4: According to Step 2 and Step 3, the coupled control module integrates the bicycle model and the tire stiffness model to generate control instructions for the steering, acceleration and speed of the driverless car to improve the vehicle's handling performance and safety and ensure the vehicle's stable driving at high speeds and under complex road conditions.

[0133] As an embodiment of the present application, the step Step 1 specifically includes:

[0134] Install various types of sensors on the vehicle, including speed sensors, angular velocity sensors, IMUs (inertial measurement units), lidars, and cameras, to ensure comprehensive acquisition of the vehicle's dynamic status information. The layout of the sensors needs to take into account the coverage and accuracy of data collection to ensure that the key parameters of the vehicle can be monitored in real time;

[0135] Real-time collection of vehicle speed, steering angle, lateral acceleration and other parameters. The vehicle speed sensor obtains the current vehicle speed by detecting changes in wheel speed, the steering angle sensor measures the steering wheel rotation angle, and the IMU provides the vehicle's acceleration and angular velocity information in all directions. LiDAR and cameras can assist in obtaining data about the surrounding environment to ensure a comprehensive understanding of the dynamic state;

[0136] The collected data is processed by filtering algorithms to eliminate noise and unnecessary interference. Common filtering methods include Kalman filtering and low-pass filtering to ensure that the dynamic state data obtained is highly accurate and real-time;

[0137] Based on the monitored dynamic state, the parameters of the unmanned driving controller are calculated and updated in real time, including control gain, preset target speed, and expected steering angle, etc. These parameters will be used for subsequent control decisions to ensure that the vehicle can maintain good handling performance in complex driving environments.

[0138] As an embodiment of the present application, the step Step 2 specifically includes:

[0139] Select a simplified model suitable for the dynamic characteristics of the driverless car, and regard the vehicle as a single body with the front wheels controlling the steering and the rear wheels controlling the drive. This simplification can effectively reduce the computational complexity while retaining the necessary dynamic characteristics, facilitating the application of subsequent control algorithms;

[0140] Determine the key parameters that affect the dynamics model, including vehicle mass, center of gravity, front and rear wheelbase, and tire characteristics. The choice of center of gravity has an important impact on the stability and handling performance of the vehicle, so it needs to be reasonably set according to the actual vehicle characteristics;

[0141] Based on the kinematics of the vehicle, the dynamic equations describing the vehicle's motion are derived. These equations usually include equations for lateral and longitudinal motion, and use Newton's second law and the equation of motion to express the relationship between the vehicle's acceleration, velocity, and displacement and control inputs (such as steering angle, throttle, and brake);

[0142] When building the dynamic model, factors such as the lateral stiffness and longitudinal stiffness of the tire are considered, and a suitable tire model is used to describe the effect of the tire on vehicle handling. This process ensures that the model can more accurately reflect the interaction between the tire and the road when turning and accelerating;

[0143] By comparing the simulation results with the measured data, the accuracy and effectiveness of the established model are verified. Necessary parameter adjustments are made to improve the applicability of the model;

[0144] Establish a real-time update mechanism to input the dynamic parameters collected in Step 1 into the dynamic model and calculate the vehicle's motion state in real time. This mechanism ensures that the model always reflects the current dynamic behavior of the vehicle and provides accurate state information for the control system;

[0145] The output of the model is integrated into the control system so that subsequent control strategies can be optimized and adjusted based on the accurate dynamic model.

[0146] As an embodiment of the present application, the step Step 3 specifically includes:

[0147] Based on the material properties and structural parameters of the tire, a tire stiffness model is established. This model takes into account the lateral stiffness and longitudinal stiffness of the tire and uses relevant mathematical equations to describe the deformation behavior of the tire when subjected to force;

[0148] Based on the vehicle state parameters collected in Step 1, calculate the tire's side slip angle when turning. The side slip angle refers to the angle between the tire's driving direction and the tire's actual driving direction. This parameter directly affects the vehicle's handling stability and turning performance;

[0149] Based on the above calculations and adjustments, a comprehensive control command is generated to guide the vehicle's steering system to perform the required steering action to ensure the stability and safety of the vehicle during the turning process.

[0150] As an embodiment of the present application, the step 4 specifically includes:

[0151] Integrate the steering control information obtained in Step 3 with the bicycle model established in Step 2 to ensure that the dynamic characteristics of the two are compatible with each other. This integration process needs to take into account the impact of tire stiffness on the dynamic behavior of the vehicle to ensure that the model can accurately reflect the actual driving situation;

[0152] Design control algorithms to compare the output of the integrated model with the set targets. Use control theory to develop control strategies and generate control commands for steering, acceleration, and speed in real time. These commands will guide the various actuators of the vehicle, such as the electric power steering system and the throttle / brake control unit;

[0153] The generated control instructions are transmitted to the vehicle's execution system, including the steering system, power system and braking system, ensuring that the vehicle can accurately execute the required dynamic behavior and achieve smooth and safe driving.

[0154] The present invention uses multiple controllers to coordinate control, and respectively performs controller coupling control on the lateral, longitudinal, tracking and obstacle avoidance of the unmanned vehicle, so that the safety of the unmanned vehicle is guaranteed, and advanced, efficient and safe unmanned driving is realized.

[0155] The unmanned driving system of this embodiment adopts multi-sensor fusion technology, integrates laser radar, camera, ultrasonic sensor and IMU, and obtains the dynamic state and surrounding environment information of the vehicle in real time. The sensor is connected to the central processing unit via the CAN bus to ensure efficient data transmission and processing. During the driving process, the vehicle monitors the vehicle speed, steering angle, lateral acceleration and surrounding obstacle information in real time. The data is processed by Kalman filtering to improve the monitoring accuracy. These dynamic parameters provide basic data for the subsequent dynamic model establishment and control instruction generation. According to the dynamic state data collected in real time, a comprehensive dynamic model considering tire stiffness and vehicle characteristics is established. This model simplifies the unmanned vehicle into a form of front-wheel control steering and rear-wheel drive to ensure accurate reflection of the dynamic behavior of the vehicle under different driving conditions. The tire stiffness model is used to calculate the tire side slip angle during steering in real time, and the steering torque is adjusted to optimize the stability of the vehicle at different speeds and road conditions. Especially when driving at high speeds, the system will increase the steering torque to ensure that the vehicle can turn safely. The coupling control module integrates the bicycle model and the tire stiffness model to generate control instructions for steering, acceleration and speed. Through PID and LQR control strategies, the vehicle can be ensured to travel smoothly, especially on complex urban roads and highways. Combined with the MPC controller, the future vehicle trajectory can be predicted and re-planned when dynamic obstacles are detected. The system dynamically adjusts the driving path by real-time evaluation of the environment and vehicle status to ensure safe avoidance of obstacles. Through the sensors installed on the roof, the system can identify up to 20 dynamic obstacles within a range of 50 meters. Combined with dynamic programming and quadratic programming algorithms, the optimal obstacle avoidance path is generated to ensure that the vehicle can safely and stably bypass obstacles. The system has a real-time feedback mechanism to continuously monitor the difference between the dynamic state of the vehicle and the control target. The control strategy is adjusted through feedback and the model parameters are optimized to improve the overall performance of the system. Through the above preferred embodiments, the unmanned vehicle system of the present invention not only has excellent handling performance and safety, but also can flexibly cope with complex driving environments, laying a solid foundation for the application of unmanned driving technology.

[0156] The above is an exemplary description of the present invention in combination with the accompanying drawings. It is obvious that the specific implementation of the present invention is not limited to the above-mentioned method. As long as such non-substantial improvements are made by adopting the method concept and technical solution of the present invention, or the concept and technical solution of the present invention are directly applied to other occasions without improvement, they are all within the protection scope of the present invention.

Claims

1. A driverless car coupling control system based on a dynamic model, characterized in that: include: The sensor module is used to obtain the driving status of the driverless car, including parameters such as vehicle speed, steering angle, and lateral acceleration; The bicycle dynamics model modeling module builds the dynamics model of the driverless car based on the parameters collected by the sensor module. The model simplifies the driverless car into a bicycle model with the front wheel controlling the steering and the rear wheel controlling the drive. The steering control module calculates the tire slip angle during steering based on the tire stiffness model and adjusts the steering torque to optimize the vehicle's turning stability and safety; The coupling control module generates control instructions for the steering, acceleration and speed of the driverless car based on the bicycle model created by the bicycle dynamics model modeling module and the tire stiffness model used by the steering control module, thereby improving the vehicle's handling performance and safety.

2. The unmanned vehicle coupling control system based on a dynamic model according to claim 1, characterized in that: The system realizes lateral control of the driverless car through a linear quadratic regulator controller, and the lateral control includes: Obtain the lateral offset and heading angle error of the vehicle based on the bicycle model; The linear quadratic regulator controller calculates the optimal steering angle in real time according to the offset and the error to achieve lateral stability and trajectory tracking of the driverless car.

3. The unmanned vehicle coupling control system based on a dynamic model according to claim 1, characterized in that: The system also includes a dual-loop PID controller for longitudinal control of the driverless vehicle, the longitudinal control including: The first PID control loop is used to control the distance between the vehicle and the target in front, and to maintain a safe distance by adjusting the throttle and brake; The second PID control loop is used to perform speed control based on the deviation between the current speed of the vehicle and the preset speed, and to achieve precise speed control by adjusting the throttle or brake.

4. The unmanned vehicle coupling control system based on a dynamic model according to claim 3, characterized in that: The dual-loop PID controller adjusts the acceleration and deceleration of the vehicle in real time to ensure the longitudinal control stability of the vehicle under complex road conditions, and dynamically adjusts the speed and distance control parameters according to traffic conditions.

5. The unmanned vehicle coupling control system based on a dynamic model according to claim 1, characterized in that: The system also implements tracking control of the driverless car through a model predictive controller, and the tracking control includes: Predict the vehicle trajectory at future moments based on the vehicle’s current state, target trajectory, and dynamics model; The model predictive controller calculates the optimal steering, acceleration, and braking operations by optimizing the control inputs to keep the vehicle along the target trajectory while satisfying the vehicle dynamics constraints and safety constraints.

6. The unmanned vehicle coupling control system based on a dynamic model according to claim 5, characterized in that: The model predictive controller can update the target trajectory according to real-time environmental changes and vehicle status, and adjust the control input to ensure vehicle tracking accuracy and safety while taking into account road curvature, obstacles and vehicle dynamics.

7. The unmanned vehicle coupling control system based on a dynamic model according to claim 1, characterized in that: The system also includes an obstacle recognition module, which realizes obstacle detection through a sensor installed on the top of the driverless car. The obstacle detection includes: The sensors detect the environment within 50 meters around the driverless car; Up to 20 targets can be identified and corresponding obstacle information can be generated based on the distance, speed and direction of the target.

8. The unmanned vehicle coupling control system based on a dynamic model according to claim 7, characterized in that: The sensors include laser radar, millimeter wave radar and camera to improve the accuracy of obstacle recognition and environmental adaptability, especially to achieve accurate detection of dynamic and static obstacles under complex road conditions.

9. The unmanned vehicle coupling control system based on a dynamic model according to claim 1, characterized in that: The system can achieve static obstacle avoidance by generating an ST graph (space-time graph), wherein the static obstacle avoidance includes: A dynamic programming algorithm is used to plan the optimal driving path based on the obstacle positions and path information in the ST diagram to ensure that the vehicle bypasses static obstacles.

10. The unmanned vehicle coupling control system based on a dynamic model according to claim 9, characterized in that: The system can avoid dynamic obstacles through a quadratic programming algorithm, and the dynamic obstacle avoidance includes: Combined with the time dimension in the ST diagram, the moving trajectory and speed of the dynamic obstacle are calculated, and a quadratic programming model is generated to adjust the vehicle's trajectory in real time to ensure that the vehicle's driving stability and safety are maintained while avoiding dynamic obstacles.

Images