Research On Motion Planning Algorithm Of Mobile Manipulator In Narrow Passage

Motion planning of mobile manipulators is a hot research direction in the field of robotics.The mobile manipulator integrates the advantages of the mobile platform and the manipulator,which not only has the mobile function of the mobile platform,but also has the flexible operation function of the manipulator.Thus,the application scope of the mobile robot is expanded.On the one hand,the mobile manipulator is a complex robot system.While performing independent control of the mobile platform and the manipulator,the coordinated control of the overall system must also be considered.On the other hand,when the robot is in a complex working environment with many obstacles or narrow passages,the use of traditional methods may cause problems such as long working hours and difficulty in finding a path.In this paper,the mobile manipulator is taken as the main object.The overall coordinated control ability and path planning ability of the mobile manipulator in a complex environment are analyzed and studied at the theoretical level.The main contents of the research are as follows:(1)The overall kinematics model of the mobile manipulator is analyzed.The mobile platform and the manipulator of the mobile manipulator are located in different workspace,which leads to the difficulty of coordination planning.From the perspective of space mapping,the various variables of the mobile platform and the manipulator are mapped to the same coordinate space for description,and the overall kinematics of the system is determined,so that the overall planning of the mobile manipulator can be unified.(2)A hierarchical planning method for mobile manipulator is proposed.Aiming at the problem that it is difficult for mobile manipulator to collect samples in narrow passage based on traditional sampling algorithms,a hybrid sampling strategy is proposed to improve the distribution of samples on the roadmap.It improves the recognition degree of difficult areas such as narrow passages by the mobile manipulator,and further improves the efficiency and passing rate of the mobile manipulator path planning in the complex environment.(3)The overall collision detection capability of the mobile manipulator is studied.In view of the long calculation time for collision detection of the whole mobile manipulator,the sampling configurations of the manipulator are projected on the two-dimensional plane,so that the configurations of the mobile manipulator can carry out collision detection in the same dimension.When the mobile platform moves towards the end point without collision,the manipulator can also avoid obstacles and maintain a safe state through trajectory planning.(4)A combined sampling strategy is proposed to make the mobile manipulator maintain stable performance when it moves in a complex environment.By combining the advantages of bridge test and PCA method,the PCA method determines the difficult area and the bridge test focuses on sampling.Through the recognition of difficult areas,the adaptive effect of sampling in different regions can be achieved.The smaller connected roadmap is generated with fewer samples,which improves the overall performance of mobile manipulator in motion planning.