Design Of Motion Planning Algorithm And Software Development Of Wheel-track-Legged Robot For Post-disaster Rescue

China has complex geographical and climatic conditions,with many types of natural disasters occurring over a wide area and causing serious damage.Mobile robots,with their small size and flexible movements,are widely used in areas such as explosion prevention and early warning and disaster rescue.Wheel-track-legged robots have various modes of locomotion such as wheeled,tracked,legged and hybrid modes.They can choose the best mode of locomotion according to the specific terrain and make full use of the advantages of wheel,track and leg structures.Therefore,the use of wheel-track-legged robots for disaster relief has good prospects for application.In the post-disaster rescue environment,there are various working conditions.When the environment is unknown,the robot needs to be continuously guided to use the on-board sensors to complete the sensory construction of the environment,and when the environment is known,the robot needs to be able to realize the motion planning under multiple working conditions(shortest rescue time or least energy consumption).Therefore,the research on motion planning of wheel-track-legged robots in post-disaster rescue has important practical significance and engineering value.Based on the national key research and development program"Wheel-track-legged autonomous unmanned vehicle development"(No.:2020YFC1511703),this paper carries out the motion planning algorithm design and software development of wheel-track-legged robot for postdisaster rescue,including:Motion planning software architecture design of wheel-track-legged robot,research on autonomous environment exploration of wheel-track-legged robot in uncertain post-disaster environment,research on global motion planning of wheel-track-legged robot in multiple working conditions,development of motion planning software of wheel-track-legged robot and physical experiment verification.First,the architecture design of the motion planning software is carried out from the perspective of engineering implementation for postdisaster rescue scenarios.Then,the overall architecture design is based on the idea of software layered architecture to determine the functional division of each module and its hierarchical structure and reduce the coupling degree between software modules;according to the requirements analysis and architecture of the motion planning software,we analyze the type and transmission direction of information of each module in the software and design the workflow and data flow between modules.The detailed analysis and design of each functional module is carried out separately.Second,considering the unknown environment,conduct research on the autonomous environment exploration problem of wheel-track-legged robots in uncertain environments after disasters.For the map description problem of uncertain environments,combined with the motion characteristics of wheel-track-legged robots,an improved occupancy grid map representation is proposed;and the autonomous exploration of unknown environments is divided into autonomous exploration in local environments and autonomous exploration in global environments;the local path planning of the exploration path is realized by using the dynamic window method;and the simulation effect is demonstrated in Gazebo.Then,considering the known environment,conduct research on the global motion planning of wheel-track-legged robots under different working conditions.Considering two typical working conditions of shortest rescue time and minimum energy consumption,use distance grid map and energy grid map to describe the environment characteristics respectively;considering the wide search area of the post-disaster rescue environment,improve the A*path planning algorithm by adding heuristic weight values to achieve a trade-off between path optimality and search speed.Finally,based on the improved A*algorithm,path planning is realized under distance constraint and energy constraint respectively and verified by simulation using Matlab.Finally,carry out the development and physical experiment verification of wheel-track-legged robot motion planning software.Based on the software architecture design,complete the development of motion planning software,and combine with hardware platform to carry out physical experiments of autonomous exploration method in uncertain environment and multi-working condition global motion planning method in known environment respectively.Further verify the feasibility and effectiveness of the theoretical methods proposed in this paper.