Research On The Path Planning And Path Following Method For Robot Based On Safe A~* And Adaptive Window Approach

With the development of the artificial intelligence,mobile robot gains it's increasing popularity in public places,such as offices,nursing homes,restaurant,airports and etc.Compared with the conventional industrial robot,mobile robots need to analyze,judge and make decisions independently and achieve autonomous navigation when they face more terrible environment.Therefore,there is no inevitably they will face complex problems such as path planning,path following and obstacle avoidance.Currently,according to the mastery of environment information,path planning algorithms can be mainly divided into global and local path planning.A single path planning algorithm is not easy to solve these problems at the same time.therefore,this paper proposes a hybrid path planning method which combined the guide point to solve navigation problem of mobile robots in complex dynamic environment.First of all,aiming at ensuring safe global path planning of a mobile robot,this paper proposes Safe A*algorithm which considers the risk cost of the map as a safety index.the improved algorithm can reduce the risk of global path by bypassing narrow passages or crowded areas.Then,Selecting the key path points from global path to remove the redundant nodes.And these key global path points are integrated to local planning when they served as a guide point.Consequentially,the complex path following are divided into multi segment local motion planning tasks.Secondly,the local motion planning and the path following control technology of mobile robot is researched.In order to improve the synthetical performance of the expectation trajectory generated by local path planning,Present AWA used in this paper analyzes deeply local environment model,in which the traversability,safety,a guidance of target,and stability of movement were taken into consideration.Simultaneously,A new velocity control law of the mobile robot based on dynamic constraints is introduced to track a destined trajectory exactly.what's more,this paper designs the strategy of switching the guide point in real time which ensures the effectiveness of the target tracking by simulating human navigation method using map.Finally,According to the results of various simulations,robustness,timeliness and environmental suitability of the above-mentioned algorithms are verified respectively.The results show high performance in terms of robustness,safety and feasibility.Besides,the experiment on the robot shows that the new method is feasible,can be able to achieve fast,safe and accurate path planning,path following and obstacle avoidance base on a given target.