A path-following output feedback control strategy for agricultural wheeled tractors

Abstract

The present invention discloses an output feedback control strategy for path tracking of an agricultural wheeled tractor, and belongs to the field of agricultural machinery navigation technology. The main steps are: 1. Establish a path tracking model containing disturbances, and convert it into a state equation in a strict feedback form; 2. Design a second-order accurate differentiator to realize the estimation of unknown states related to heading in the state equation and lumped disturbances; 3. Based on the goal of path tracking, select a sliding surface and establish a first-order sliding mode dynamics equation; 4. Design an output feedback controller to achieve the path tracking goal. Advantages of the present invention: 1. The controller design only uses position deviation information, and achieves the goal of path tracking without using sensors to measure heading deviation information, thereby reducing sensor costs; 2. The controller shortens the system response time and improves tracking accuracy; 3. The disturbance in the system is accurately estimated and synchronously compensated to the controller, thereby enhancing the system's anti-disturbance performance.

Description

Technical Field

[0001] The present invention relates to a path tracking control technology for a navigation system of an agricultural wheeled tractor, specifically, a path tracking algorithm with output feedback is designed by using state observation technology and a super spiral control method. The invention aims to improve the transient performance, tracking accuracy and stability of the navigation system of an agricultural wheeled tractor, and belongs to the field of agricultural machinery navigation technology. Background Art

[0002] The mechanization and automation of agricultural production is the basis for the implementation of precision agriculture. The higher the degree of mechanization and automation of agricultural production, the more conducive it is to the implementation of precision agriculture technology. As an important power source for field mechanized operations, agricultural tractors can be used with various agricultural machinery to achieve a series of field operations and field management tasks, and can also be used to tow trailers to achieve transportation operations. However, the control effect of the automatic navigation system is affected by factors such as the mechanical structure of the agricultural machinery itself, posture sensors, operating conditions, and control algorithms. This results in the automatic navigation system of agricultural tractors not being able to achieve satisfactory tracking effects in actual work. In order to solve this problem, the path tracking control algorithm of the agricultural machinery navigation system is studied, so that the automatic navigation system can not only ensure the operation accuracy, but also improve the robustness of the navigation system to cope with disturbances.

[0003] Generally, the design of the path tracking controller of the agricultural tractor navigation system requires the use of vehicle position information and heading information. However, since the sensors installed on the agricultural machinery for measuring the heading are susceptible to measurement noise and vehicle shaking, this will cause large errors in the measured value of the heading information, resulting in abnormal fluctuations in the control signal, affecting the path tracking effect. Therefore, the present invention intends to use state observation technology to simultaneously observe the unknown system state related to the heading in real time and estimate the unknown lumped disturbance in real time; on this basis, the superhelical control method is used for output feedback control design. It is worth noting that the developed path tracking control strategy only needs to measure the vehicle's position information, and does not use sensors to measure the heading information to achieve vehicle control; in addition, the path tracking control algorithm is effective in improving transient path tracking performance, eliminating steady-state errors, improving stability, and suppressing disturbance directions. Summary of the invention

[0004] This paper proposes an output feedback control strategy for path tracking of an agricultural wheeled tractor, which achieves the path tracking control target without using the heading information measured by the sensor. The designed path tracking control algorithm is effective in improving the transient path tracking performance of the tractor navigation system, eliminating steady-state errors, improving stability, and suppressing disturbances.

[0005] An output feedback control strategy for path tracking of an agricultural wheeled tractor. The specific design steps are as follows:

[0006] Step 1: Analyze the disturbance factors existing in the actual operation process of the agricultural wheeled tractor, build a path tracking model including the disturbance, and use it as a reference model for the design of the path controller;

[0007] Step 2: introduce coordinate transformation to transform the path tracking model obtained in step 1 into a state equation in a strict feedback form that is convenient for controller design;

[0008] Step 3: Design a second-order accurate differentiator for the unknown system state and unknown lumped disturbance in the state equation obtained in step 2, and simultaneously realize real-time observation of the unknown state and accurate estimation of the unknown lumped disturbance;

[0009] Step 4: Based on the path tracking goal, select appropriate sliding variables, combine the state equation obtained in step 2 and the exact differentiator in step 3, and establish the first-order sliding mode dynamic equation;

[0010] Step 5: Based on the first-order sliding mode dynamics equation constructed in step 4, an output feedback super-helical controller based on state observation technology is designed, and then, through inverse transformation, the actual control input of the vehicle's front wheel steering angle is obtained.

[0011] Specifically, in step 1, the path tracking model of the agricultural wheeled tractor is established as follows:

[0012]

[0013] Among them, l os and θ os They represent lateral deviation and heading deviation respectively, σ is the directional coefficient, v is the longitudinal velocity, l t is the vehicle wheelbase, δ is the front wheel steering angle, c d is the curvature of the reference path, and d(t) is the lumped disturbance including system uncertainty and external interference.

[0014] In step 2, in order to facilitate controller design, the path tracking model (1) is converted into a standard form of strict feedback that is convenient for controller design. For convenience, it is assumed that the vehicle moves forward and tracks the reference path in a clockwise manner, that is, the direction coefficient σ = -1, v> 0. Let x 1 = l os , x 2 =v sinθ os , u = tanδ, then system (1) can be re-expressed as follows:

[0015]

[0016] Among them, x 1 and x 2is the state variable and u is the virtual control input.

[0017] In step 3, for the unknown system state x in system (2), 2 , design a second-order accurate differentiator for the unknown state x 2 To facilitate the design of the state observer, the system (2) is re-expressed as follows:

[0018]

[0019] in, is the lumped disturbance; for the unknown system state x in system (3) 2 and the unknown lumped disturbance Δ, the second-order accurate differentiator is designed as follows:

[0020]

[0021] Among them, L 1 , L 2 and L 3 is a positive real observation gain, and Represents the observer output variable. By choosing a reasonable disturbance observation gain, the output state Used to observe unknown state x in real time 2 , output status Used to estimate the lumped disturbance Δ.

[0022] In step 4, the goal of agricultural tractor path tracking is to make the lateral deviation and heading deviation converge to zero; to this end, the observation value of the unknown state obtained in (4) is used. Select the sliding variables as follows:

[0023]

[0024] where ζ is a positive control parameter. Further, by taking the derivative of the sliding variable s along the system (3) and combining it with the state observer (4), we can obtain the first-order sliding mode dynamics equation:

[0025]

[0026] in, They are state x 1 and x 2 The observation error.

[0027] In step 5, the output feedback continuous superhelical controller u based on the state observation technology is designed as:

[0028]

[0029] where k 1 >0,k 2 > 0, the sliding variable s will be stable in a finite time.

[0030] Furthermore, the virtual controller u = tan(δ) is inversely transformed, and the actual front wheel steering angle δ is:

[0031]

[0032] The lateral deviation l os and heading deviation θ os converges to zero.

[0033] The outstanding beneficial effects of the present invention are as follows:

[0034] 1. In the present invention, the path tracking model is converted into a state equation in a strict feedback form for control design, which reduces the difficulty of controller design; the controller design only uses the measured position deviation information, and realizes the control of the vehicle without using sensors to measure heading deviation information, thereby reducing the sensor cost.

[0035] 2. The path tracking algorithm in the present invention can not only ensure that the lateral deviation and heading deviation converge to zero within a limited time, but also obtain faster system response and higher tracking accuracy.

[0036] 3. The method of the present invention is simple and easy to implement. It estimates the disturbance in the system through a second-order precise differentiator and compensates it synchronously in the controller, thereby enhancing the system's robustness against disturbances and having a better control effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a control block diagram of the path tracking control system of the agricultural wheeled tractor of the present invention.

[0038] Figure 2 It is a schematic diagram of the path tracking of the agricultural wheeled tractor of the present invention.

[0039] Figure 3 is the curve of the disturbance d(t) changing with time.

[0040] Figure 4 It is the response curve of lateral deviation under U-shaped path condition.

[0041] Figure 5 It is the response curve of heading deviation under U-shaped path condition.

[0042] Figure 6 It is the response curve of the front wheel steering angle under U-shaped path condition.

[0043] Figure 7 It is the trajectory result of path tracking under U-shaped path condition. DETAILED DESCRIPTION

[0044] The present invention provides an output feedback control strategy for path tracking of an agricultural wheeled tractor. In order to make the purpose, technical solution and effect of the present invention clearer and more specific, the technical solution in the embodiment of the present invention will be described clearly and completely in conjunction with the accompanying drawings in the embodiment of the present invention. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0045] Figure 1 Shown is a control block diagram of the agricultural wheeled tractor path tracking control system of the present invention, which mainly includes a control loop module and an observation loop. Figure 2 The schematic diagram of the path tracking of the agricultural tractor of the present invention. The reference speed v of the vehicle is 3m / s, and the wheelbase of the vehicle is l t is 1.69m, and the control parameter ζ is 0.5; in addition, it is assumed that the vehicle tracks the reference path in a clockwise manner, that is, the direction coefficient σ=-1. An output feedback control strategy for path tracking of an agricultural wheeled tractor, the implementation process of the method is:

[0046] Step 1: Establish the path tracking model of the agricultural wheeled tractor including disturbance as follows:

[0047]

[0048] Among them, l os and θ os They represent lateral deviation and heading deviation respectively, σ is the directional coefficient, v is the longitudinal velocity, l t is the vehicle wheelbase, δ is the front wheel steering angle, c d is the curvature of the reference path, and d(t) is the lumped disturbance including system uncertainty and external interference.

[0049] Step 2: Convert the path tracking model (1) into a state equation in strict feedback form.

[0050] Let x 1 = l os , x 2 =v sinθo s , u = tanδ, then the system model (1) is written as the state equation as follows:

[0051]

[0052] Among them, x 1 and x 2 is the system state and u is the virtual control input.

[0053] System (2) can be reformulated as follows:

[0054]

[0055] in, is the lumped disturbance.

[0056] Step 3: Design a second-order accurate differentiator as follows:

[0057]

[0058] Among them, L 1 , L 2 and L 3 is a positive real observation gain, and represents the observer output variable, where the output state and They are used to quickly observe the unknown state x 2 and the lumped disturbance Δ.

[0059] Step 4: Controller design of superhelical output feedback based on state observation technology.

[0060] The goal of agricultural tractor path tracking is to make Figure 2 The lateral deviation in os and heading deviation θ os Converges to zero, for this purpose, the sliding variable is selected as follows:

[0061]

[0062] where ζ is a positive definite control parameter.

[0063] Furthermore, by taking the first-order derivative of the sliding variable s along the system (3) and combining it with the state observer (4), the first-order sliding mode dynamic equation can be obtained as follows:

[0064]

[0065] in, They are state x 1 and x 2 The observation error.

[0066] According to the first-order sliding mode dynamics equation obtained above, the output feedback continuous superhelical controller u based on state observation technology is designed as:

[0067]

[0068] where k 1 >0,k 2 > 0, the sliding variable s will be stable in a finite time.

[0069] Furthermore, the virtual controller u = tan(δ) is inversely transformed, and the actual front wheel steering angle δ is:

[0070]

[0071] The lateral deviation l os and heading deviation θ os converges to zero.

[0072] In order to better verify the control effect of the proposed output feedback path tracking algorithm, the present invention built a simulation platform based on Matlab software to verify the effectiveness of the controller in the presence of interference. The simulation uses the Euler method and sets the sampling period to 0.001ms.

[0073] Figure 3 is the curve of disturbance d(t) changing with time, Figure 4 is the curve of lateral deviation changing with time under U-shaped path condition, Figure 5 is the response curve of the heading deviation under the U-shaped path condition, Figure 6 is the response curve of the front wheel steering angle under the U-shaped path condition, Figure 7 It is the trajectory result of path tracking under U-shaped path condition.

[0074] It can be seen from the simulation results that in the presence of interference, the superhelical output feedback controller designed by the present invention based on the state observation technology can track the reference path and travel along the reference path in a relatively short time. At the same time, the proposed control algorithm has good robust performance.

[0075] Although the present invention has been described according to various specific embodiments, those skilled in the art will appreciate that the present invention can be implemented with modifications within the spirit of the claims. Therefore, any obvious improvements, substitutions or modifications that can be made by those skilled in the art without departing from the essence of the present invention belong to the protection scope of the present invention.

Claims

1. An output feedback control strategy for path tracking of an agricultural wheeled tractor, characterized in that: The following steps are involved: Step 1: Analyze the disturbance factors existing in the actual operation process of agricultural wheeled tractors. Considering that agricultural machinery will be affected by unmodeled dynamics, side slip effects, and unknown external disturbances in actual working scenarios, a path tracking model containing disturbances is constructed. It is used as a reference model for the design of the path controller. The established path tracking model containing disturbances is as follows: Among them, l os and θ os They represent lateral deviation and heading deviation respectively, σ is the directional coefficient, v is the longitudinal velocity, l t is the vehicle wheelbase, δ is the front wheel steering angle, c d is the curvature of the reference path, d(t) is the lumped disturbance, which mainly includes unmodeled dynamics, sideslip effects, and unknown external disturbances; Step 2: Introduce coordinate transformation to transform the path tracking model obtained in step 1 into a state equation in a strict feedback form that is convenient for controller design, which is more conducive to control design. The specific implementation process is as follows: Assume that the vehicle is considered to be moving forward, and the vehicle follows the reference path in a clockwise manner, that is, the direction coefficient σ = -1; let x1 = l os ,x2=υsinθ os , u = tanδ, then system (1) can be re-expressed as follows: Among them, x1 and x2 are system states, and u is a virtual control input. To facilitate the processing of the unknown system state x2, the system (2) is further expressed as follows: in, represents the unknown lumped disturbance; Step 3: Design a second-order accurate differentiator for the unknown system state and unknown lumped disturbance in the state equation obtained in step 2, and realize real-time observation of the unknown state and accurate estimation of the unknown lumped disturbance. The construction form of the second-order accurate differentiator is as follows: Among them, L1, L2 and L3 are positive real observation gains, and represents the observer output variable; it is worth noting that the output state in observer (4) and They are used to quickly observe the unknown state x2 and the unknown lumped disturbance Δ respectively; Step 4: Based on the goal of path tracking, the unknown state is estimated by combining the state equation obtained in step 2 and the precise differentiator in step 3 using the second-order precise differentiator. Select appropriate sliding variables and establish the first-order sliding mode dynamics equation as follows: Based on the path tracking goal of the agricultural wheeled tractor, the sliding variables are selected as follows: Among them, ζ is a positive control parameter; then, the first-order derivative of the sliding variable s along the system (3) is calculated, and combined with the state observer (4), the first-order sliding mode dynamic equation is obtained: in, denote the observation errors of system states x1 and x2 respectively; Step 5: Based on the first-order sliding mode dynamics equation constructed in step 4, an output feedback super-helical controller based on state observation technology is designed, and then the actual control input of the vehicle front wheel steering angle is obtained through inverse transformation. The output feedback super-helical controller u based on state observation technology is designed as: Among them, k1>0, k2>0, then the sliding variable s will be stable in a finite time; Furthermore, by performing an inverse transformation on the virtual controller u=tan(δ), the front wheel steering angle δ is obtained as: The lateral deviation l os and heading deviation θ os converges to zero.

Images