Design And Implementation Of Extreme Driving Path Assistance Planning System For Ugv In Complicated Unstructured Environment

In order to meet the increasingly wide demand for the application of unmanned ground vehicles,more and more unmanned ground vehicles are applied for the complicated unstructured environment,which is characterized by unobvious road boundaries and various environmental impact factors,and thus the unmanned ground vehicle auxiliary planning system in the traditional structured environment is facing new challenges.Since the path planning in the complicated unstructured environment is affected by various factors such as environmental texture,terrain and slope,it's hard to perfectly tell the differences between obstacle area and passable area when driving in such environment.The ultimate driving ability of the unmanned ground vehicles poses huge impacts on the driving status in the field environment while the decision makers,usually in need of more accurate and comprehensive auxiliary information,give assistance to do path selections.Based on the above-mentioned analysis on demand,taking the auxiliary planning of unmanned ground vehicles for the complicated unstructured environment as the study object,and in consideration of the ultimate operating conditions restraints,this thesis designs and realizes an ultimate driving auxiliary planning system for the complicated unstructured environment.Besides,by modeling analysis on the complicated unstructured environmental impacts on the planning and searching for the paths in the environmental space to form some available paths,this thesis provides a comprehensive path cost evaluation method so as to contribute to developing the auxiliary ability of the decision makers on the path planning.First and foremost,the auxiliary planning system in this thesis has done the abstraction and modeling on the complicated unstructured environment.Map rasterization,slope presentation,obstacle judgment and classification of environmental textures help to achieve the analysis on the factors affecting path planning.This thesis makes full use of the contour line extraction to describe the undulation of terrain,and the corner constraint and slope constraint of unmanned ground vehicles are taken into account at the same time.This is how the spatial modeling and cost mapping method of solving the ultimate-operating-condition driving of unmanned ground vehicles in the complicated unstructured environment take shape.Secondly,the auxiliary planning system achieves the path against the environmental constraints and the ultimate operating conditions of unmanned ground vehicles,and at the same time,provides a comprehensive path cost evaluation method.The multiple search method based on Monte Carlo random sampling which has been applied to the path planning provides a heuristic solution to the path planning.Additionally,cost evaluation on the path has been made from the perspectives of safety,stability,path length,fuel consumption and steering respectively,and a coefficient weighted comprehensive cost evaluation method is provided,to assist the decision makers in selecting a "optimal" path in line with the task requirements at present.In the end,this thesis establishes an unmanned ground vehicle auxiliary planning verification system in the simulation environment,and achieves the path auxiliary planning of unmanned ground vehicles for the complicated unstructured environment.The analysis on the rationality of the planning results presents the verification on the feasibility and effectiveness of auxiliary planning system design proposed in this thesis.