Research On Path Planning Method Of Lunar Rover

Since the first cruising remote-controlled robot "Lunar patrol 1" landed on the moon,the development of planetary rover capable of autonomously detecting extraterrestrial bodies such as the moon and Mars has become one of the hot issues in the aerospace field.Path planning of planetary rover is crucial for the autonomous operation,safe roaming and detection.At present,most research on path planning focuses on terrain passability,and rarely involves potential constraints such as illumination and communication.Human decision-making may help the rover to deal with emergencies that the autonomous navigation fails to work and thus improve the performance of the entire path planning system.In this dissertation,the environmental constraints,the motion constraints of rover and the man-machine collaboration mechanism are comprehensively integrated into the existing path planning algorithm to solve its insufficient environmental adaptability and limited ability to respond to emergencies.The complex lunar environment and the motion characteristics of the lunar rover will affect the performance of path planning.In this dissertation,the study on modelling path planning constraints is implemented via considering these factors.In view of the lunar surface detection scene,a mathematical model describing the terrain,illumination and communication is established.Starting from the motion characteristics of the lunar rover,three typical motion states including unit uphill,downhill and obstacle crossing are taken into account to establish the rover’s motion constraint model.After that,an improved A* path planning algorithm is developed via integrated the lunar surface environment constraints.Based on the Digital Elevation Model(DEM)of the lunar surface released by the official website of the US Geological Survey,the dangerous area of the lunar DEM is marked in combination with the angles of the sun and the earth,and the motion parameters of the rover to provide a basis for path planning.Through the weighted fusion of environmental factors such as terrain,illumination,communication,etc.,the cost estimation function of the improved A* algorithm is formulated.Numerical simulations via MATLAB software show that the path planned by the improved A* algorithm is more suitable for lunar surface detection since dangerous areas,such as steep slopes,light shadows,and invisible communications,can be effectively avoided.On this basis,all activities of the rover are divided into six behaviors:movement,perception,detection,charging,data transmission,and dormancy.The time consumption model of the each behavior is established to complete the global path planning experiment under the changing environmental constraints.Then,due to the low resolution of the existing lunar surface DEM,it is difficult to truly reflect the lunar surface terrain,such as lunar rocks,craters,and other small obstacles which affect the rover’s normal locomotion.In this context,an improved D*local path planning algorithm is prosed to ensures that the rover moves safely and reliably in scenes with unknown obstacles.The core idea of this method is to integrate these factors,such as lunar surface environmental constraints,rover steering angle,obstacle distance and so on,into the cost function of D* algorithm.Using the simulated lunar surface environment established in Unity,the performance of the improved path planning algorithm is verified via MATLAB and GAZEBO simulation experiments.It may be concluded that the improved D* algorithm is feasible and local path planning is necessary for the lunar’s safety.Finally,a path planning system architecture based on man-machine collaboration is introduced and a GUI software is developed to realize man-machine interaction.The feasibility and effectiveness of the man-machine collaboration path planning system are verified through ground mobile trolley experiments.