On-line Deformable Terrain Parameter Identification Based On Wheel-terrain Interaction Mechanics

For the driving conditions of off-road vehicles in deformable terrain, Bekker’s semi-empirical model of terrain mechanics can be used to define the relationship between the force of the traveling mechanism(wheels, tracks) and terrain with a set of terrain mechanics parameters. Obviously, Bekker’s terrain mechanics model reveals that the terrain mechanics parameters will affect the force performance of the traveling mechanism. Thereby it will also affect the operation state and rideability of the vehicle. Therefore, the accurate identification of the terrain parameter is the basement of off-road vehicle’s performance optimization and control.When the traditional off-road vehicles(foot, wheeled, crawler)travel or work in the soft area, poor trafficability, low efficiency, high energy consumption and poor adaptability are the problems that need to conquer. This paper combines the wheelled structure with the air cushion and develops the prototype. The wheeled air-cushion off-road vehicles can regulate the driving force and the lift force depending on ground conditions, and then change the wheels contact pressure. Thereby the kinematics and dynamics, the adaptability, the application robustness of vehicle will be improved. Meanwhile, vertical force-degree-of-freedom control is also beneficial for decoupling terrain mechanics model and solving the multiple solutions of terrain mechanics parameters identification.Firstly, this paper has analyzed the function system of the wheeled air cushion off-road vehicles to determine the overall structure of the prototype model. Then the project design of the air cushin system, the drive control system, the power supply system are completed which involves the modeling, the type selection and the development. And the test of the fan performance and the test of the efficiency of inverter are proceed.Then, according to the force/ torque integral equations between the soft terrain and the wheel, an equivalent decoupling method and Romberg integration method are proposed. After that the two decoupled analytical model are derived and then accurate comparative study is conducted.Next, according to the equivalent decoupling method, three shear parameters and two pressure parameters can be identified online step by step for the air cushion off-road vehicles and the traditional off-road vehicles. The improved Newton algorithm has been proposed to identify the three shear parameters using the measured values of the vertical load, the slip rate, the traction and the driving torque.Afterwords, the identification result of three shear parameters will serve as the known conditions to identify the two pressure parameters using BFS algorithm. This paper studies the successive influence of the accuracy of stepwise valuation.Finally, this paper proposes three methods to improve the accuracy of identification. Im Newton algorithm and BFS algorithm has solved the problem of multiple solutions, and we use mean calculation, sample rearrangement and other methods to improve the accuracy of the identification algorithms.Our research in this paper is funded by the National Natural Science Foundation Project "based on deliberative / reactive hybrid control architecture of air off-road soft ground autonomous robot navigation research"(61203351).