Research On Slip Behavior Of Mars Rover For On-Orbit Application

Deep space exploration is highly involved in China’s space activities,which is of great and far-reaching significance for the sustainable development of China’s politics,economy,science and society.With the steady development of China’s deep space exploration missions,higher requirements are set for the safety and autonomy of planetary vehicles patrolling and detecting in the unknown and unstructured environment of the star surface.The slippage problem is a key problem that affects the safe driving of planetary rovers.Almost all planetary rovers that have successfully fulfilled star surface inspection missions are faced slippage hazards.High-slip events will result in the sinking of planetary rover.Low-medium slip events will lead to deviation of the rover from the planned path and increase the risk of collision with geometric obstacles.Taking the Zhurong Mars rover as its research object,this paper,focusing on the sliding behavior of the rover,attempts to establish a complete research system for slip estimation,slip path planning,and slip trajectory tracking.The research results obtained so far include:(1)In view of the fact that Mars rover’s on-orbit gravitational acceleration and star soil conditions is different from that on the Earth,a sliding ground test is conducted in a test site equipped with a low-gravity simulation device and simulated star soil,which is similar with the working conditions on the surface of Mars.Based on the ground test,a total of 86,997 slip observation vector data was collected,the centralized trend,discrete trend and risk trend of the data were comprehensively analyzed,the longitudinal slip behavior and lateral slip behavior of the rover were compared and analyzed,the impact of wheel speed,star soil and slope on slip behavior was analyzed,and finally a prediction model of slip risk assessment quantity HSF-slope was established;(2)In light of the fact that the quality of the Mars rover system is strictly limited so that it is impossible to equip redundant sensors for real-time slip estimation,this paper puts forward a technical solution for slip estimation using limited resources on the device.Based on the ground test data obtained through Back Propagation network(BP),GA-BP,and Gaussian Process Regression(GPR)model,the appropriate signal is selected from the existing sensor signals to form the input feature vector,the vertical slip and lateral slip are used to form the output feature vector,multiple sets of slip estimation models are established,and then the models are evaluated based on the ground validation set data.Subsequently,based on the real telemetry data of the "Zhurong" Mars rover in orbit,the feasibility of on-orbit use of multiple sets of models was analyzed and verified.The verification results show that,GA-BP model and GPR model have good on-orbit applicability,and the uncertainty confidence intervals of the slip estimated value can be obtained through the GPR model,which will provide basic values for subsequent slip path planning.(3)In view of the fact that the working environment of the Mars rover is highly unstructured and with high slip uncertainty,this paper establishes an accessibility map construction method considering slip factors,and proposes a global path planning algorithm considering slip uncertainty.First,the Mars surface environment modeling was carried out to determine the mathematical model of the Mars surface characteristics(rocks,impact craters and crescent dunes)and their distribution density functions,and to establish the global layer and local layer Mars surface environment.Combined with the slip-slope model established above,an accessibility map that comprehensively considers geometric obstacles and non-geometric obstacles is established.Based on the slip uncertainty proposed above,this paper proposes a path planning algorithm that considers the slip uncertainty,which avoids the collision between the planetary vehicle and the geometric obstacle caused by the accumulation of slip uncertainty with the extension of the path.(4)In light of the fact that the rover runs very slowly(nominal operating condition is 200m/h)and the slip behavior will affect the trajectory tracking accuracy,this paper proposes,based on the slip kinematics model,a model prediction trajectory tracking control method,established a slip kinematics model based on differential steering strategy and two-way Ackerman steering strategy.Through simulation analysis,the trajectory tracking effects of different arc radii,different steering strategies,different slip angles,and different arc combinations are compared,which provides a strategic basis for the Mars rover on-orbit control.