Terramechanics-based analysis and control for lunar/planetary exploration robots

Abstract

Surface mobility by using wheeled mobile robots (Rovers) is one of the important technologies for lunar/planetary exploration missions. The rover in these missions significantly expands the exploration area and thus increases the scientific or programmatic return from the mission. These planetary exploration rovers are expected to travel long distances and perform complex tasks in order to fulfill challenging mission goals. Surface terrain of the Moon or a planet such as Mars is covered with fine-grained soil called regolith, or boulders, rocks, or stones spread over the terrain. Because of the challenging terrain, the rover should be aware of mobility hazards, such as wheel slip/stuck, vehicle tip-over, and collision with obstacles. This research addresses mainly three issues: (1) development of an analytical model for investigations of the motion behavior of the rover on loose soil, (2) path following control with considering the slip-compensated approach, and (3) path planning and performance evaluation method to obtain the safest path in order to tackle the slip dynamic problem in the path planning issue. For the issue of the development of an analytical model for motion behavior, this research deals with two models. First, a wheel-soil contact model are developed based on Terramechanics approach in order to address the wheel slip/skid motion. Subsequently, the motion behavior of the rover is numerically obtained by using a wheel-and-vehicle model. The wheel-soil contact model can calculate three wheel forces and two torques generated around the wheel. The validity of the model is confirmed by the experiment using a single-wheel test bed, and also, the experimental results are compared with numerical simulation results using the wheel-soil contact model. The wheel-and-vehicle model is elaborated upon by incorporating the wheel-soil contact model into an articulated multibody model of the vehicle. To analyze the motion behavior of the rover, dynamics simulation using the proposed model is conducted. In the simulation, contact forces for all wheels are computed by the wheel-soil contact model. Then, positions, orientations, and velocities of the rover can be numerically obtained by solving the forward dynamics of the vehicle. The wheel-and-vehicle model is verified through the comparison between steering experiments and corresponding simulations using the proposed model. Also, steering maneuver analysis of the rover on loose soil is addressed. Using the proposed analytical model for motion behavior, the mobility and trafficability analyses are presented in this research. The slope climbing/traversing capabilities of the rover is discussed based on the terramechanics. Here, two criteria for the slope traversability, Mobility limit and Trafficability limit, are introduced. Further, the thrust-cornering characteristics curve is developed to analyze the slope traversing capability. As the second issue of this research, the path following control taking wheel slippages is described by applying our background in terms of dynamic slip motion. In this dissertation, two control approaches are described: sensor-based feedback control and model-based feedforward control. The former approach is to calculate a steering maneuver based on the thrust-cornering characteristics curve in advance, by considering the force equilibriums between traction load of vehicle and wheel traction forces. On the other hand, the latter approach is conducted based on a feedback control using distance and orientation errors, which are measured by sensors mounted on the rover. The performance of the path following strategy is confirmed through both numerical simulations and actual slope traversal experiments. In the final issue of this research, the path-planning algorithm and the performance evaluation are described. The proposed approach for this issue is based on the following approach: First, a path on a rough terrain is generated with the terrain-based criteria function. Subsequently, the dynamics simulation of a rover is carried out in which the rover is controlled to follow the candidate path. Finally, the path is properly evaluated based on the simulation results, such as slip motion profiles, vehicle orientations, and energy consumptions. Demonstrations for the proposed technique are addressed along with a discussion on characteristics of the candidate path and the dynamic profile of the rover. The techniques developed and proposed in this work: (1) analytical model for motion behavior of rover (2) path following control with slip compensation (3) path planning and performance evaluation would be beneficial for the application of not only lunar/planetary exploration rovers but also field mobile robots on rough challenging terrain.