Research On Motion Control Methods For Asymmetric Structural Lunar Rover With Six Wheels And Its Parameters Identification

Lunar rover, as a pioneer to the lunar surface exploration, is mainly able to complete complex tasks of the probing and investigating of the lunar surface and the collecting and analyzing of the samples. The reasonable design of the mechanical structure and the excellent motion control are the keys to ensure the success of the lunar exploration. The fast development of computer technology, communication technology and automatic control technology provides a broader space for the exploitation and the researches of the lunar rover. The motion control of the lunar rover, which is the key technology, is becoming the focus of research and development. This paper presents a new kind of lunar rover structure with freedom knowledge property right, which is six wheels differential steering and asymmetrical structure, and studys the technology of kinematics modeling, trajectory tracking, driving control, and the motion parameter identification in depth.Through the analysis and comparison of the existing lunar rover's(planetary vehicles) structural features, this paper proposes a new statically indeterminate asymmetric six-wheeled lunar rover configuration and builds two types of kinematics models based on the features of the rugged and flat terrain. The six wheels were asymmetrically distributed at both sides of the vehicle body. And the rocker arms, which could lift up freely, are connected with the body by torsion springs. The non-symmetrical distribution makes it easier to cross the obstacles such as the wide and deep ditches with diameters longer than that of the wheels. By active control mode the rocker arms can still offer traction when the wheels skid, which improves the ability of crossing the rugged terrain. According to the specific structural features of the lunar rover, the kinematical model was built with the use of the multi-body system principle. And the Matlab simulation has shown the model is correct. Moreover, based on the analysis of the lunar rover's plane motion features, this paper builds a kinematic model of the flat terrain and analysizes the wheel-slipping phenomenon (longitudinal slip and lateral slip) in detail.In regard to the control problem of the lunar rover's kinematic model, satisfying the nonholonomic constraints or not satisfying nonholonomic constraints, two tracking control laws are designed based on sliding mode control. For the control problem of satisfying the nonholonomic constraints without considering wheel slip, the sliding mode trajectory tracking control law with global asymptotic stability, is designed by choosing the switch function reasonably and adopting backstepping technology. This law should be based on the existence and stability of sliding mode. For the control problem without satisfying the nonholonomic constraints but considering wheel slip, the sliding mode trajectory tracking control law satisfying the Lyapunov stability and finite time convergence is designed. This law is built by using diffeomorphism principle to transform the kinematic model to a perturbed chain system, and by defining a new time varying sliding surface. Theoretical analysis and simulation results show that the two control laws are effective and feasible. The lunar rover trajectory tracking is finally achieved by using the laws.To solve the drive control of the rover in the complex lunar terrain, this paper puts forward a variable structure control algorithm, which is based on the improved genetic algorithm PID tuning parameters of sliding surface. Considering the population diversity and the fitness of calibration in the traditional genetic algorithm, the method of defining the similarity function of R and the calibration function of fitness is adopted in order to increase the population diversity and prevent local optimum. This method has effectively improved the global stability and optimizing speed of the genetic algorithm. Based the above, a sliding mode controller is designed by using the improved genetic algorithm to tune the parameters of PID sliding surface. With the help of the sliding mode controller, the uncertainty of controlling caused by external disturbances has been solved. The simulations and experiments shows that the method is stable and feasible.In order to control and guide the lunar rover accurately with wheel-terrain contact angle and slip value estimated in real time under the condition of complex wheel-terrain interactions and kinematics constraints, the mathematical model of the lunar rover's visual navigation is analyzed firstly, and a kinematics equation concerning slip and contact angle is established. Then, a close loop is formed with the drive and guidance systems, the data about motion path of the lunar rover's centroid is measured, and a method is presented to identify slip values and wheel-tarrain contact angle with the help of the observer. The method can ensure accurate and safe motion of lunar rover, and increase control precision. Simulation and experimental result demonstrate the effectiveness and feasibility of the method.The software testing procedure of the ground test system is developed. To the structural features of the testing lunar rover sample, its basic performances of movements are measured through the experiments of the crossing of ditches and obstacles, climbing and the folding of wheel-leg. To the emerging problems of the experiment ( such as wheel slipping, traction dissatisfied ), two experiments have been done by using the theories of the chapters above. One experiment is the tracking experiment with coordination of the wheels, which is based on the parameter identification technique. The other one is the self-recovered traction experiment, which is based on the prediction of the traction. The problems emerged are finally solved comparatively well through the experiments.