Research On Adaptive Navigation In Dynamic Scene Of Indoor Mobile Robot Combined With Learning Mechanism

Along with sharp increase in labor cost,and people's pursuit for higher quality in life and better efficiency in work in the recent years,indoor mobile robot,as one representation of mobile robot,played an important role on individuals' daily life and work.However,navigation is precondition for indoor mobile robots performing other tasks;Due to specificity of indoor environment,such as narrow space and irregular-shape obstacles,indoor environment has higher requirements for navigation's real-time performance and safety.Hence,traditional methods in mobile robot navigation are no longer fairly suitable when it comes to indoor environment.Focusing on the specificity of indoor environment,this paper proposes research on adaptive navigation in dynamic scene of indoor mobile robot combined with learning mechanism.The main research work is as follows:Firstly,history of indoor mobile robot is introduced.Popular mobile robot navigation methods are presented,and for each,we conclude their advantages and disadvantage.Then,the outline of this paper's work is described.After that,an IHDR algorithm is introduced and applied to global navigation of indoor mobile robot.Besides,we make detailed descriptions of IHDR and analyze its pros and cons.Secondly,an escaping from obstacle zone method for mobile robot based on maximum distance priority mechanism is proposed.This method perceives local environment by distance information,which has a good performance in real-time and safety.Furthermore,a maximum distance priority mechanism is innovatively devised in this method to overcome robot endless wandering.Thirdly,a method of adaptive navigation in dynamic scene of indoor mobile robot is put forward,which combines the learning-based navigation method with the escaping from obstacle zone method,and make mobile robot suit the indoor environment well.Finally,we apply the proposed methods in MT-R robot,and results of experiments illustrate feasibility and effectiveness of the method.Meanwhile,indoor robot navigation control software used in the experiment is introduced.This paper provides a good reference for indoor mobile robot navigation and relevant field.