Improved LQR transverse control method and system for heavy commercial vehicle

Abstract

The invention discloses an improved LQR transverse control system for a heavy commercial vehicle. The improved LQR transverse control system comprises a sensing preview module, a data processing module, an LQR control module and a steering execution module. The sensing preview module is used for sensing road information and sending the road information to the data processing module; the data processing module is electrically connected with the sensing preview module, and the sensing preview module is used for carrying out data processing on the road information and sending the processed data to the LQR control module; the LQR control module is electrically connected with the data processing module, and the LQR control module is used for calculating a transverse distance error and a steering angle error of the vehicle and sending calculation results to the steering execution module; and the steering execution module is electrically connected with the LQR control module, and the steering linear module is used for receiving the calculation result and controlling vehicle steering according to the calculation result.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of heavy commercial vehicle lateral control, and particularly relates to a heavy commercial vehicle improved LQR lateral control method and system. BACKGROUND

[0002] With the upgrading of vehicle intelligent technology, automatic driving technology has become the key to improving transportation efficiency and safety, especially suitable for heavy commercial vehicles, especially tractor vehicles, to quickly land in the logistics transportation industry. As a core module of automatic driving, lateral control needs to ensure that the vehicle is stable in lane centering control or path tracking under various working conditions in high-speed logistics transportation scenarios. Therefore, a lateral control method suitable for heavy commercial vehicles is urgently needed to solve the problems of control precision and accuracy caused by semi-trailer control, large load change control, etc.

[0003] Traditional lateral control methods such as Pure Pursuit, fuzzy control, MPC, etc. simple kinematic model or single control strategy have been widely used in the field of passenger cars, but for commercial vehicles, especially semi-trailer tractors, there are problems such as unsuitable semi-trailer dynamics model, large load change range, and insufficient real-time detection accuracy, etc. Direct adaptation will face great limitations.

[0004] The prior art scheme detects the current lateral deviation through a perception system, calculates the error of the target lateral deviation according to path planning, uses a PID algorithm to eliminate the lateral deviation error to realize path tracking, or establishes a two-degree-of-freedom dynamics model, uses a quadratic linear regulator LQR to calculate a reward function, and outputs a steering wheel angle or front wheel angle control quantity to a steering machine for direction control.

[0005] When the prior art is applied to commercial vehicles, the following problems exist:

[0006] (1) Large control error: the single control method of LQR has a large lateral control error in the continuous curve scene of heavy commercial vehicles.

[0007] (2) Overshoot or insufficient response caused by large load change; the control parameters of PID lateral control cannot be calibrated for most loads due to the large load change range of commercial vehicles, and the real-time detection accuracy of load values cannot be guaranteed for MPC and LQR, which can easily lead to overshoot or insufficient response of the output control quantity.

[0008] (3) Not suitable for semi-trailer tractors: the existing two-degree-of-freedom dynamics model is used for LQR lateral control of semi-trailer tractors, and the control accuracy is low due to the fact that the dynamics model does not conform to reality. SUMMARY

[0009] The application aims to provide an improved LQR lateral control method and system for heavy commercial vehicles, which combines multi-point preview technology to reduce lateral error in curved road conditions, and uses a variable center of gravity LQR control algorithm to solve the control problem of a semi-trailer tractor, can input accurate load to reduce control overshoot or insufficient response, and the improved LQR lateral control algorithm is more suitable for heavy commercial vehicles, including heavy trucks and semi-trailer tractors.

[0010] In order to solve the above problems existing in the prior art, the technical scheme adopted by the application is:

[0011] An improved LQR lateral control system for heavy commercial vehicles, comprising a perception preview module, a data processing module, an LQR control module and a steering execution module.

[0012] The perception preview module is used for sensing road information and sending the road information to the data processing module.

[0013] The data processing module is electrically connected with the perception preview module, and the perception preview module is used for data processing of the road information and sending the processed data to the LQR control module.

[0014] The LQR control module is electrically connected with the data processing module, and the LQR control module is used for calculating vehicle lateral distance error and steering angle error and sending the calculation results to the steering execution module.

[0015] The steering execution module is electrically connected with the LQR control module, and the steering straight line module is used for receiving the calculation results and controlling vehicle steering according to the calculation results.

[0016] An improved LQR lateral control method for heavy commercial vehicles, comprising the following steps:

[0017] Step 1: detecting the road lane information in front, setting the preview position according to the current vehicle speed v; setting five preview positions D1, D2, D3, D4 and D5 according to the current vehicle speed, wherein DX=a*v, and a is a calibration coefficient.

[0018] Step 2: obtaining lateral distance error and steering angle error according to the preview position; in the step 2, the lateral distance and steering angle error are obtained according to the preview position and the path planning reference track line.

[0019] Step 3: data processing of the lateral distance error and the steering angle error to obtain the mean value of the lateral distance error and the mean value of the steering angle error; in the step 3, Kalman filtering is performed on the lateral distance and the heading angle error output by the perception preview module to avoid data jump caused by frame loss and the like, and then the mean value of the five lateral distances and the mean value of the five heading angle errors are output.

[0020] Step 4: Calculate the control amount δ in combination with the vehicle information, and output the control amount δ;

[0021] The step 4 includes the following steps:

[0022] Step 41: Determine the linear state space model of the vehicle Establish a vehicle model;

[0023] The vehicle model includes a heavy truck model and a tractor-semitrailer model. The heavy truck model is:

[0024]

[0025] The tractor-semitrailer model needs to be extended in state variables based on the heavy truck model:

[0026]

[0027] Where θ A is the articulation angle e tractor between the tractor and the trailer, e trailer is the lateral deviation of the tractor, θ trailer is the lateral deviation of the trailer, θ tractor is the heading angle of the trailer, θ yf is the heading angle of the tractor, F yr is the front / rear wheel lateral force, M coupling is the articulation point coupling moment, l A is the distance from the trailer mass center to the articulation point.

[0028] Assuming that the trailer mass center is on the articulation point, the extended tractor-semitrailer model is:

[0029]

[0030] Where m tractor is the mass of the tractor, m tralier is the mass of the trailer, l tr is the distance from the trailer articulation point to the rear axle, l t is the wheelbase of the tractor, K θA is the articulation angle torsional stiffness, D θA is the articulation angle damping, I tr is the trailer rotational inertia.

[0031] Step 42: Select the state weight matrix Q and the control input weight matrix R to solve the quadratic cost function

[0032] Solve the feedback gain K to obtain the control amount δ = -Kx.

[0033] x is the system state vector, u is the control input vector, in the above vehicle dynamics model, in order to avoid the real-time detection accuracy of the load weight, the load weight of the vehicle is input by the driver, according to the accurate cargo quality input by the driver, the m, l f 、l r (Truck), m trailer 、l A (Tractor trailer) and other values, so that the heavy commercial vehicle dynamics model is more accurate, the control amount overshoot or response is reduced, and the lateral control stability is enhanced. In order to obtain accurate cargo weight, if no input is selected, the real-time detection estimated value is selected. According to the change of the cargo weight, the current vehicle head and the mass of the trailer are combined, and the m, l f 、l r (Truck), m trailer 、l A (Tractor trailer) and other values, so that the LQR control is more accurate.

[0034] Step 43: output the control amount delta.

[0035] Step 5: receive the control amount delta, and execute according to the control amount delta.

[0036] The beneficial effects of the present application are:

[0037] (1) Control accuracy: the five-point preview method reduces the lateral control error in the curve scene.

[0038] (2) Control stability: accurate load weight input is more suitable for control of heavy commercial vehicles, reduces control amount overshoot or response deficiency, and enhances lateral control stability.

[0039] (3) Strong applicability: the dynamics model extended by using the state variables of the tractor trailer improves the practicability of the lateral control method and meets the control requirements of heavy vehicles and tractor trailers.

[0040] In summary, the present application provides a heavy commercial vehicle improved LQR lateral control method and system, which has the advantages of high control accuracy, good control stability and applicability to heavy tractor trailers, and provides an accurate and stable lateral control algorithm for high-speed automatic driving in the logistics industry. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 The flowchart of the present application is shown. DETAILED DESCRIPTION

[0042] The present application will be further described below in combination with the drawings and reference numerals.

[0043] In order to enable the above-mentioned objects, features and advantages of the present application to be clearer, the present application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0044] The terms "first", "second", "third", etc. are only used for differentiation and cannot be understood as indicating or implying relative importance.

[0045] In the description of the present application, it should also be noted that, unless otherwise explicitly specified and limited, the terms "arrange", "mount", "connect", "connect" should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0046] The specific embodiments of the present application will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application.

[0047] Embodiment 1

[0048] An improved LQR lateral control system for a heavy commercial vehicle, comprising: a perception preview module, a data processing module, an LQR control module, a steering execution module;

[0049] The perception preview module is used to perceive road information and send the road information to the data processing module;

[0050] The data processing module is electrically connected with the perception preview module, and the perception preview module is used to process the road information data and send the processed data to the LQR control module;

[0051] The LQR control module is electrically connected with the data processing module, and the LQR control module is used to calculate the vehicle lateral distance error and the steering angle error, and send the calculation results to the steering execution module;

[0052] The steering execution module is electrically connected with the LQR control module, and the steering execution module is used to receive the calculation results and control the vehicle steering according to the calculation results.

[0053] Embodiment 2

[0054] As shown in Figure 1 An improved LQR lateral control method for a heavy commercial vehicle, comprising the following steps:

[0055] Step 1: Detect the front road lane information, and set the preview position according to the current vehicle speed v;

[0056] Step 2: Obtain the lateral distance error and steering angle error according to the preview position;

[0057] Step 3: Process the lateral distance error and steering angle error data to obtain the average lateral distance error and average steering angle error;

[0058] Step 4: Calculate the control amount δ in combination with vehicle information, and output the control amount δ;

[0059] Step 5: Receive the control amount δ and execute according to the control amount δ.

[0060] Embodiment 3:

[0061] In the step 1 of the embodiment 2, five preview positions D1, D2, D3, D4 and D5 are set according to the current vehicle speed, where DX=a*v, and a is a calibration coefficient.

[0062] Embodiment 4:

[0063] In the step 2 of the embodiment 3, the lateral distance error and steering angle error are obtained according to the preview position and the path planning reference trajectory line.

[0064] Embodiment 5:

[0065] In the step 3 of the embodiment 4, Kalman filtering is performed on the lateral distance error and steering angle error output by the perception preview module to avoid data jumping caused by frame loss and the like, and then the average of the five lateral distances and the average of the five steering angles are output.

[0066] Embodiment 6:

[0067] In the embodiment 5, the step 4 includes the following steps:

[0068] Step 41: Determine the linear state space model of the vehicle Establish a vehicle model;

[0069] The vehicle model includes a heavy truck model:

[0070]

[0071] The vehicle model includes a semi-trailer tractor model, which extends the state variables based on the heavy truck model:

[0072]

[0073] Where θ A is the hinge angle between the tractor and the trailer, etrailer is the trailer lateral deviation, trailer is the trailer heading angle,

[0074] F yf ,F yr is the front / rear wheel side force, M coupling is the hinge point coupling moment, l A is the trailer mass center to hinge point distance.

[0075] The extended model is

[0076]

[0077] where C f ,C r is the front / rear tire cornering stiffness, l f ,l r is the mass center to front / rear axle distance, m tractor is the host vehicle mass, m tralier is the trailer mass, I z is the moment of inertia, v x is the longitudinal vehicle speed, l tr is the trailer hinge point to rear axle distance, l t is the tractor wheel base, K θA is the hinge angle torsional stiffness, D θA hinge angle damping, I tr is the trailer moment of inertia.

[0078] In the above vehicle dynamics model, in order to avoid real-time detection accuracy of the load weight cannot be guaranteed, the vehicle load uses the driver input strategy to obtain accurate cargo weight, if no input, select the real-time detection estimated value. According to the change of cargo weight, combined with the current vehicle head and trailer mass, mass center, real-time calculation and input dynamics model parameters m, l f ,l r (load truck), m trailer ,l A (semi-trailer tractor) and other values, so that the LQR control is more accurate.

[0079] Step 42: select state weight matrix Q, and control input weight matrix R, solve quadratic cost function

[0080] Solve the feedback gain K, get the control amount δ = -Kx.

[0081] Step 43: output control amount δ.

[0082] The present application is not limited to the above-mentioned optional embodiments, and anyone can derive other various forms of products under the inspiration of the present application, but regardless of any changes in shape or structure, any technical solutions falling within the scope defined by the claims of the present application fall within the protection scope of the present application.

Claims

1. An improved LQR lateral control system for heavy-duty commercial vehicles, characterized in that, include: Perception and anticipation module, data processing module, LQR control module, steering execution module; The perception and aiming module is used to perceive road information and send the road information to the data processing module; The data processing module is electrically connected to the perception and pre-aiming module. The perception and pre-aiming module is used to process road information and send the processed data to the LQR control module. The LQR control module is electrically connected to the data processing module. The LQR control module is used to calculate the vehicle's lateral distance error and steering angle error, and sends the calculation results to the steering execution module. The steering execution module is electrically connected to the LQR control module, and the steering line module is used to receive the calculation results and control the vehicle steering according to the calculation results.

2. An improved LQR lateral control method for heavy-duty commercial vehicles, characterized in that, Includes the following steps: Step 1: Detect lane information ahead and set the aiming position based on the current vehicle speed v; Step 2: Based on the pre-aiming position, obtain the lateral distance error and steering angle error; Step 3: Process the data for lateral distance error and steering angle error to obtain the mean lateral distance error and the mean steering angle error; Step 4: Calculate the control quantity δ based on the vehicle information, and output the control quantity δ. Step 5: Receive the control quantity δ and execute according to the control quantity δ.

3. The improved LQR lateral control method for heavy-duty commercial vehicles according to claim 2, characterized in that, In step 1, five aiming positions D1, D2, D3, D4, and D5 are set according to the current vehicle speed, where DX = a * v, and a is a calibration coefficient.

4. The improved LQR lateral control method for heavy-duty commercial vehicles according to claim 2, characterized in that, In step 2, the lateral distance and steering angle error are obtained based on the pre-aiming position and the path planning reference trajectory line.

5. The improved LQR lateral control method for heavy-duty commercial vehicles according to claim 2, characterized in that, In step 3, the lateral distance and heading angle error output by the perception and aiming module are subjected to Kalman filtering to avoid data jumps caused by missing frames and other reasons. Then, the average of the five lateral distances and the average of the five heading angle errors are output.

6. The improved LQR lateral control method for heavy-duty commercial vehicles according to claim 2, characterized in that, Step 4 includes the following steps: Step 41: Determine the linear state-space model of the vehicle Establish a heavy-duty truck model or a semi-trailer tractor model; Step 42: Select the state weight matrix Q and the control input weight matrix R, and solve for the quadratic cost function. Solving for the feedback gain K, we obtain the control quantity δ = -Kx. Step 43: Output control quantity δ.

7. The improved LQR lateral control method for heavy-duty commercial vehicles according to claim 6, characterized in that: In step 41, the heavy-duty truck model is as follows: Where C f C r For the lateral stiffness of the front / rear tires, l f l r I is the distance from the center of gravity to the front / rear axle, m is the vehicle mass, and I is the distance from the center of gravity to the front / rear axle. z v is the moment of inertia. x This refers to the longitudinal vehicle speed.

8. The improved LQR lateral control method for heavy-duty commercial vehicles according to claim 6, characterized in that: In step 41, the semi-trailer tractor model is as follows: Where C f C r For the lateral stiffness of the front / rear tires, l f l r The distance from the center of mass to the front / rear axle, in meters. tractor Main vehicle mass, m tralier For trailer weight, I z v is the moment of inertia. x For longitudinal vehicle speed, l tr l is the distance from the trailer articulation point to the rear axle. t K represents the wheelbase of the tractor unit. θA For the hinge angle torsional stiffness, D θA Hinged angle damping, I tr This represents the moment of inertia of the trailer.

Images