Cross-country road vehicle motion control method and device based on MPC

Abstract

The invention provides an MPC-based cross-country road vehicle motion control method and device, and partially solves the technical problems of poor safety and stability during high-speed driving on a cross-country road. The method comprises the following steps: modeling according to physical characteristics of the cross-country terrain to form a terrain dynamics model for identifying and quantifying terrain constraints; performing terrain constraint in the path planning process of planning the vehicle dynamics model in the sensing environment through model prediction control to form real-time optimization control of a motion track; and a kinematics control instruction of a vehicle low-level controller is formed according to the real-time optimization control to adjust the motion trail of the vehicle so as to keep stability and safety. A reliable solution is provided for an autonomous driving system passing through a cross-country road surface at a high speed, and the method has important practical application prospects.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle control, and in particular to an MPC-based off-road vehicle motion control method and device. Background Art

[0002] In the existing technology, the autonomous driving technology for high-speed off-road roads attempts to use physical modeling, model predictive control and other technologies to achieve rapid calculation and constraint processing of terrain dynamics. For example, terrain constraints are identified based on a physical framework and key constraints for safe driving are derived, and learning dynamic models are used to improve driving safety. It also includes attempts to apply advanced control algorithms and sensing technologies to improve the safety and stability of vehicles in these environments. However, in terms of safety and stability, there are still high technical challenges in terms of rapid calculation of constraint processing, accuracy of terrain constraint identification, effectiveness of path planning, and accuracy of dynamic models. Summary of the invention

[0003] In view of the above problems, an embodiment of the present invention provides an off-road vehicle motion control method and device based on MPC, which partially solves the technical problems of poor safety and stability when driving at high speed on off-road roads.

[0004] The MPC-based off-road vehicle motion control method of the embodiment of the present invention includes:

[0005] Modeling the physical characteristics of off-road terrain to form a terrain dynamics model to identify and quantify terrain constraints;

[0006] Through model predictive control, the vehicle dynamics model is planned to perform real-time optimization control of the motion trajectory in the process of path planning in the perceived environment under terrain constraints;

[0007] The kinematic control instructions of the vehicle's low-level controller are formed according to real-time optimization control to adjust the vehicle's motion trajectory to maintain stability and safety.

[0008] In one embodiment of the present invention, the modeling of the terrain dynamics model based on the physical characteristics of the off-road terrain for identifying and quantifying the terrain constraints includes:

[0009] Classify the terrain according to different geological structural characteristics and form corresponding terrain dynamics models;

[0010] The dynamic characteristics of the vehicle under different terrain conditions are analyzed according to the terrain dynamics model, and the corresponding terrain constraint conditions are established.

[0011] In one embodiment of the present invention, the real-time optimization control of forming a motion trajectory by terrain constraints in the process of path planning of the vehicle dynamics model in the perceived environment through model predictive control includes:

[0012] Collect environmental information through comprehensive sensors to form spatial data of perceived environment;

[0013] Forming a vehicle dynamics model based on terrain constraints;

[0014] The model predictive control process is used to form the path planning of the vehicle dynamics model in the perception environment, and the optimized control data of the real-time state of the vehicle dynamics model is formed according to the path motion trajectory.

[0015] In one embodiment of the present invention, the kinematic control instructions of the vehicle low-level controller are formed according to the real-time optimization control to adjust the motion trajectory of the vehicle to maintain stability and safety, including:

[0016] The optimized control data is converted into kinematic control instructions for the low-level controller to form the actual control of the vehicle.

[0017] The MPC-based off-road vehicle motion control device according to an embodiment of the present invention comprises:

[0018] A memory for storing program codes in the process of processing the above-mentioned MPC-based off-road vehicle motion control method;

[0019] A processor is used to execute the program code.

[0020] The off-road vehicle motion control device based on MPC in the embodiment of the present invention comprises:

[0021] A physical modeling module, which is used to model the physical characteristics of the off-road terrain to form a terrain dynamics model, and is used to identify and quantify terrain constraints;

[0022] The planning control module is used to perform real-time optimization control of the motion trajectory formed by terrain constraints in the process of path planning of the vehicle dynamics model in the perceived environment through model predictive control;

[0023] The real-time response module is used to adjust the vehicle's motion trajectory to maintain stability and safety based on the kinematic control instructions of the vehicle's low-level controller formed by real-time optimization control.

[0024] In one embodiment of the present invention, the physical modeling module includes:

[0025] A terrain modeling unit is used to classify terrain according to different geological structural characteristics and form a corresponding terrain dynamic model;

[0026] The constraint quantification unit is used to analyze the dynamic characteristics of the vehicle under different terrain conditions according to the terrain dynamics model and establish corresponding terrain constraint conditions.

[0027] In one embodiment of the present invention, the planning control module includes:

[0028] A data acquisition unit, used to collect environmental information through integrated sensors to form spatial data of perceived environment;

[0029] a model adaptation unit for forming a vehicle dynamics model according to terrain constraints;

[0030] The path planning unit is used to form the path planning of the vehicle dynamics model in the perception environment by using the model predictive control process, and to form the optimized control data of the real-time state of the vehicle dynamics model according to the path motion trajectory.

[0031] In one embodiment of the present invention, the real-time response module includes:

[0032] The control implementation unit is used to convert the optimized control data into kinematic control instructions for the low-level controller to form the actual control of the vehicle.

[0033] The MPC-based off-road vehicle motion control method and device of the embodiment of the present invention achieves comprehensive optimization of the autonomous driving system that passes through off-road surfaces at high speeds by combining physical modeling, model predictive control and low-level controller response. It can effectively improve the safety and stability of the vehicle on off-road surfaces. Through physical modeling and constraint processing, the system can more accurately identify and respond to terrain constraints to avoid dangerous situations such as rollovers and crossing gullies. At the same time, the combination of the MPC algorithm and the low-level controller enables the vehicle to achieve precise control of the driving trajectory, maintain stability during high-speed driving, and improve driving comfort and safety. Overall, this technical solution provides a reliable solution for autonomous driving systems that pass through off-road surfaces at high speeds, and has important practical application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 FIG. 1 is a flow chart of an off-road vehicle motion control method based on MPC according to an embodiment of the present invention.

[0035] Figure 2 FIG. 1 is a schematic diagram of the architecture of an off-road vehicle motion control device based on MPC according to an embodiment of the present invention. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solution and advantages of the present invention clearer and more understandable, the present invention is further described below in conjunction with the accompanying drawings and specific implementation methods. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] An off-road vehicle motion control method based on MPC according to an embodiment of the present invention is as follows Figure 1 shown.

[0038] exist Figure 1 In this embodiment, the present invention includes:

[0039] Step 100: Modeling the terrain dynamics model based on the physical characteristics of the off-road terrain is used to identify and quantify terrain constraints.

[0040] The types of off-road terrain include single terrain and composite terrain. Composite terrain is formed by the rapid switching or superposition of different single terrains. Different types of off-road terrain are affected by geological composition and are closely related to the vehicle's driving performance and driving stability in terms of terrain description dimensions such as rigid bearing characteristics, friction change laws, impact deformation characteristics, and surface creep laws. A terrain dynamics model is formed based on the physical characteristics of the geological composition of the off-road terrain to quantify the off-road terrain characteristics. The dynamic characteristics of the vehicle under different terrain conditions are analyzed based on the off-road terrain characteristics to form a quantitative constraint relationship between various off-road terrains and various driving states of the vehicle. The constraint relationship includes but is not limited to the stable conditions and extreme value conditions for the vehicle's state changes such as speed changes, angle changes, friction changes, and kinetic potential energy conversion for specific terrain types. Obtain terrain constraint data when the vehicle is driving on the corresponding terrain to avoid extreme scenarios such as rollovers and failure to cross gullies.

[0041] Step 200: Using model predictive control to plan the vehicle dynamics model in the process of path planning in the perceived environment, real-time optimization control of the motion trajectory is performed based on terrain constraints.

[0042] Those skilled in the art can understand that the vehicle dynamics model can describe the mechanical characteristics of the vehicle, such as the motion moment and torque. Environmental perception provides a description of the driving space and obstacles in the space. Terrain constraints constrain the motion characteristics of the vehicle dynamics model in the perceived environment. On the basis of reflecting the expected driving state through the vehicle dynamics model, a motion trajectory planning is formed in the objective environment space. In the motion trajectory planning, the terrain constraints are used by the model predictive control (MPC) to form real-time optimization and adjustment of the motion trajectory planning process of the vehicle dynamics model, so that the planned motion trajectory can more effectively respond to emergencies and terrain changes.

[0043] Step 300: The kinematic control instructions of the vehicle's low-level controller are formed according to the real-time optimization control to adjust the vehicle's motion trajectory to maintain stability and safety.

[0044] Real-time optimization control forms adjustment control data for the motion trajectory, which is converted into kinematic control instructions acting on the low-level controller. The real-time response of the low-level controller is used to make real-time optimization of the driving trajectory.

[0045] The MPC-based off-road vehicle motion control method of the embodiment of the present invention achieves comprehensive optimization of the autonomous driving system that passes through off-road roads at high speed by combining physical modeling, model predictive control and low-level controller response. It can effectively improve the safety and stability of the vehicle on off-road roads. Through physical modeling and constraint processing, the system can more accurately identify and respond to terrain constraints to avoid dangerous situations such as rollover and crossing gullies. At the same time, the combination of the MPC algorithm and the low-level controller enables the vehicle to achieve precise control of the driving trajectory, maintain stability during high-speed driving, and improve driving comfort and safety. Overall, this technical solution provides a reliable solution for autonomous driving systems that pass through off-road roads at high speed, and has important practical application prospects.

[0046] like Figure 1 As shown, in one embodiment of the present invention, step 100 includes:

[0047] Step 110: Classify the terrain according to different geological structure characteristics and form a corresponding terrain dynamic model.

[0048] The terrain dynamics model is formed based on the interaction between the terrain structural characteristics determined by terrain classification and the vehicle motion state, so that the force state between vehicle motion and terrain interaction can be quantified.

[0049] Step 120: Analyze the dynamic characteristics of the vehicle under different terrain conditions according to the terrain dynamics model and establish corresponding terrain constraint conditions.

[0050] Terrain constraints are used to establish the critical conditions for the interaction between the vehicle and the terrain when the vehicle reaches the highest speed while maintaining motion stability. On the basis of satisfying terrain constraints, the vehicle motion has driving stability and safety at high speeds.

[0051] The MPC-based off-road vehicle motion control method of an embodiment of the present invention establishes terrain constraints to quantify the motion state envelope area that maintains stability and safety when the vehicle is in a violent motion state, thereby avoiding dangerous states such as vehicle rollover, failure to climb over obstacles, and failure to cross gullies.

[0052] like Figure 1 As shown, in one embodiment of the present invention, step 200 includes:

[0053] Step 210: Collect environmental information through comprehensive sensors to form spatial data of the perceived environment.

[0054] Comprehensive sensors include but are not limited to LiDAR, RGB-D cameras and other sensors. LiDAR sensors can provide high-precision terrain information, while RGB-D cameras can capture the location of obstacles and other key information in the environment. The vehicle perceives the spatial information of the surrounding environment in real time and can use general spatial modeling technology to form a quantitative perception environment.

[0055] Step 220: Form a vehicle dynamics model according to the terrain constraints.

[0056] The terrain constraint data formed according to the terrain dynamics model is mapped to the vehicle dynamics model, so that the force state and power output state of the vehicle form an envelope range of the safety situation.

[0057] Step 230: Utilize the model predictive control process to form a path planning of the vehicle dynamics model in the perception environment, and form optimized control data of the real-time state of the vehicle dynamics model according to the path motion trajectory.

[0058] The force changes and motion control states in path planning can be quantified through the dynamic model of the vehicle under different terrain conditions. In one embodiment of the present invention, in order to improve the computational efficiency, the MPPI (model predicted path integral) control is sampled in the planar kinematic space. The expansion of the planar kinematic MPPI is projected onto the elevation map to obtain the state sequence in three-dimensional space. This includes modeling factors such as terrain features, slopes, obstacles, etc., so that these terrain constraints can be taken into account in path planning and motion control. The algorithm flow is as follows:

[0059] a. Sampling control sequence

[0060] In sample-based model predictive control (MPC), the control sequence is first sampled from a Gaussian distribution. The specific steps are as follows:

[0061] Set the control time range H.

[0062] Randomly sample N open-loop control sequences {ut:t+H-1i}i=1N.

[0063] Compute the cumulative cost for each control sequence:

[0064] C(ut:t+H-1i;xt):=∑h=0H-1c(xt+h,ut+h)

[0065] where c(xt+h,ut+h) is the cost function at time t+h.

[0066] b. Processing control sequences

[0067] In the sampling control sequence, it is necessary to ensure the feasibility of control. The specific steps are as follows:

[0068] Set the change limits of the control sequence to ensure that the changes in linear speed and curvature are within the vehicle's throttle, brake and steering limits.

[0069] Process the original control sequence {u~t:t+H-1i}i=1N to generate a feasible control sequence {ut:t+H-1i}i=1N:

[0070] vt+hi=clip(v~t+hi,vt+h-1i±Δvmax)

[0071] vt+hi is the speed at time t+h

[0072] κt+hi={κt+h-1iclip(κ~t+hi,κt+h-1i±Δκmax)if|∣vt+hi∣ <vminotherwise

[0073] κt+hi is the curvature at time t+h

[0074] c. Evaluation trajectory

[0075] The elevation map is used to evaluate the traversability of the sampled trajectory. The state sequence in three-dimensional space is obtained by projecting the expansion of the planar kinematics onto the elevation map.

[0076] d. Calculate attitude angle

[0077] Calculate the vehicle's attitude angles (roll angle and pitch angle) using the elevation map:

[0078] φ=roll(h) and θ=pitch(h)

[0079] Where h is the height of the corresponding position in the elevation map.

[0080] e. Constraints and cost functions

[0081] In the MPPI framework, traversability constraints are used as part of the cost function to avoid the risk of the vehicle rolling over or getting stuck in a ditch on uneven terrain. The cost function is in the form of:

[0082] c(x,u):=w1crollover(x,u)+w2cairtime(x,u)+w3cbump(x,u)

[0083] crollover(x,u) is the rollover cost item

[0084] cairtime(x,u) is the vehicle hanging cost item

[0085] cbump(x,u) is the collision cost term

[0086] Among them w1, w2, w3 are manually adjusted weights.

[0087] f. Update of control commands

[0088] Update the control commands using the Model Predictive Path Integral (MPPI) control algorithm:

[0089] ut = argminuC(u; xt)

[0090] The MPC-based off-road vehicle motion control method of the embodiment of the present invention uses model predictive control to combine terrain constraints in the off-road vehicle path planning process, and simultaneously forms parallel calculation of the planned path, so as to achieve a rapid response to high-speed off-road driving and ensure the stability and reliability of control.

[0091] like Figure 1 As shown, in one embodiment of the present invention, step 300 includes:

[0092] Step 310: Convert the optimized control data into kinematic control instructions of the low-level controller to form actual control of the vehicle.

[0093] The MPC-based off-road vehicle motion control method of the embodiment of the present invention can achieve real-time and precise control of the vehicle through the synergy of the MPC algorithm and the low-level controller, ensuring the safety and stability of the vehicle when passing through different terrains at high speed.

[0094] Invention Embodiment 1 An off-road vehicle motion control device based on MPC comprises:

[0095] A memory, used to store program codes in the process of processing the off-road vehicle motion control method based on MPC in the above-mentioned embodiment;

[0096] The processor is used to execute the program code in the processing of the off-road vehicle motion control method based on MPC in the above embodiment.

[0097] The processor may be a DSP (Digital Signal Processor) digital signal processor, an FPGA (Field-Programmable Gate Array) field programmable gate array, an MCU (Microcontroller Unit) system board, a SoC (system on a chip) system board, or a PLC (Programmable Logic Controller) minimum system including I / O.

[0098] Invention Embodiment 1 MPC-based off-road vehicle motion control device Figure 2 As shown. Figure 2 In this embodiment, the present invention includes:

[0099] A physical modeling module 10, for modeling a terrain dynamics model according to the physical characteristics of the off-road terrain, for identifying and quantifying terrain constraints;

[0100] A planning control module 20, for performing real-time optimization control of a motion trajectory formed by terrain constraints in a path planning process of a planning vehicle dynamics model in a perception environment through model predictive control;

[0101] The real-time response module 30 is used to adjust the motion trajectory of the vehicle to maintain stability and safety according to the kinematic control instructions of the vehicle's low-level controller formed by the real-time optimization control.

[0102] like Figure 2 As shown, in one embodiment of the present invention, the physical modeling module 10 includes:

[0103] A terrain modeling unit 11 is used to classify terrain according to different geological structure characteristics and form a corresponding terrain dynamic model;

[0104] The constraint quantification unit 12 is used to analyze the dynamic characteristics of the vehicle under different terrain conditions according to the terrain dynamics model and establish corresponding terrain constraint conditions.

[0105] like Figure 2 As shown, in one embodiment of the present invention, the planning control module 20 includes:

[0106] A data collection unit 21 is used to collect environmental information through comprehensive sensors to form spatial data of the perceived environment;

[0107] A model adaptation unit 22, for forming a vehicle dynamics model according to terrain constraints;

[0108] The path planning unit 23 is used to form a path planning of the vehicle dynamics model in the perception environment by using the model prediction control process, and to form optimized control data of the real-time state of the vehicle dynamics model according to the path motion trajectory.

[0109] like Figure 2 As shown, in one embodiment of the present invention, the real-time response module 30 includes:

[0110] The control implementation unit 31 is used to convert the optimized control data into kinematic control instructions of the low-level controller to form actual control of the vehicle.

[0111] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for off-road vehicle motion control based on MPC, characterized in that: include: Modeling the physical characteristics of off-road terrain to form a terrain dynamics model to identify and quantify terrain constraints; Through model predictive control, the vehicle dynamics model is planned to perform real-time optimization control of the motion trajectory in the process of path planning in the perceived environment under terrain constraints; The kinematic control instructions of the vehicle's low-level controller are formed according to real-time optimization control to adjust the vehicle's motion trajectory to maintain stability and safety.

2. The off-road vehicle motion control method based on MPC as claimed in claim 1, characterized in that: The modeling of the terrain dynamics model based on the physical characteristics of the off-road terrain for identifying and quantifying terrain constraints includes: Classify the terrain according to different geological structural characteristics and form corresponding terrain dynamics models; The dynamic characteristics of the vehicle under different terrain conditions are analyzed according to the terrain dynamics model, and the corresponding terrain constraint conditions are established.

3. The off-road vehicle motion control method based on MPC as claimed in claim 1, characterized in that: The real-time optimization control of forming a motion trajectory by using the model predictive control to plan the vehicle dynamics model in the process of path planning in the perceived environment under terrain constraints includes: Collect environmental information through comprehensive sensors to form spatial data of perceived environment; Forming a vehicle dynamics model based on terrain constraints; The model predictive control process is used to form the path planning of the vehicle dynamics model in the perception environment, and the optimized control data of the real-time state of the vehicle dynamics model is formed according to the path motion trajectory.

4. The off-road vehicle motion control method based on MPC as claimed in claim 1, characterized in that: The method of adjusting the motion trajectory of the vehicle to maintain stability and safety by forming a kinematic control instruction of a low-level controller of the vehicle according to real-time optimization control includes: The optimized control data is converted into kinematic control instructions for the low-level controller to form the actual control of the vehicle.

5. An off-road vehicle motion control device based on MPC, characterized in that: include: A memory for storing program codes in the process of processing the MPC-based off-road vehicle motion control method according to any one of claims 1 to 4; A processor is used to execute the program code.

6. An off-road vehicle motion control device based on MPC, characterized in that: include: A physical modeling module, which is used to model the physical characteristics of the off-road terrain to form a terrain dynamics model, and is used to identify and quantify terrain constraints; The planning control module is used to perform real-time optimization control of the motion trajectory formed by terrain constraints in the path planning process of the planning vehicle dynamics model in the perceived environment through model predictive control; The real-time response module is used to adjust the vehicle's motion trajectory to maintain stability and safety based on the kinematic control instructions of the vehicle's low-level controller formed by real-time optimization control.

7. The MPC-based off-road vehicle motion control device according to claim 6, characterized in that: The physical modeling module includes: A terrain modeling unit is used to classify terrain according to different geological structural characteristics and form a corresponding terrain dynamic model; The constraint quantification unit is used to analyze the dynamic characteristics of the vehicle under different terrain conditions according to the terrain dynamics model and establish corresponding terrain constraint conditions.

8. The MPC-based off-road vehicle motion control device according to claim 6, characterized in that: The planning control module includes: A data acquisition unit, used to collect environmental information through integrated sensors to form spatial data of perceived environment; a model adaptation unit for forming a vehicle dynamics model according to terrain constraints; The path planning unit is used to form the path planning of the vehicle dynamics model in the perception environment by using the model predictive control process, and to form the optimized control data of the real-time state of the vehicle dynamics model according to the path motion trajectory.

9. The off-road vehicle motion control device based on MPC as claimed in claim 6, characterized in that: The real-time response module includes: The control implementation unit is used to convert the optimized control data into kinematic control instructions for the low-level controller to form the actual control of the vehicle.

Images