| With the development of society and the increase of user's requirement,robots need to accomplish more and more complex tasks and provide various services for human beings.However,in the face of increasingly complex family tasks,the effect of the robot's execution is very unstable under the influence of various factors,and unpredictable problems often break out.It is likely to cause damage to expensive hardware of robot with great risk and instability.Therefore,it is very necessary to carry out virtual experiments for the robot in the simulation platform,and simulate the possible situations that may occur in the real environment.Then we can find and solve the possible problems in the actual operation of the robot,and ensure the security and accuracy of the robot that perform the task in the real environment.In this paper,we design the simulation scene and construct the 3D motion simulation platform for home service robot based on the actual situation of service robot laboratory in the Shandong University.As a complex system for simulating the motion of robot,the simulation platform for home service not only contains the necessary family item models and robot model,but also has other functional modules,such as scene editing,collision detection,path planning,motion planning,task planning and so on.They also play an indispensable role.Therefore,this paper first makes a general design of the system architecture for the simulation platform from a macro point of view,clarifies the effect of each function module and the relationship between each other,and focuses on the research of path planning module and the motion planning module.As an important part of the home service simulation platform,path planning module directly affects the robot's navigational ability in complex family environment.The family scene is not as structured as factory,and the distribution of objects is chaotic and disordered,so that the robot needs to avoid obstacles when it moves.The traditional grid method is difficult to meet the requirements of map processing and path planning in the three-dimensional complex scene.Therefore,this paper uses the A*algorithm based on navigation grid to extract the walking area in the scene map,then find the optimal path to reach the target,which greatly simplifies the amount of data needed to be processed by the path finding algorithm.Motion planning module is the key point of the simulation platform for home service,which need to be designed and researched for.As a complex mechanism with multiple joints,the motion of robot is affected by many factors and variables.Based on the fundamental kinematics analysis,the relationship between joint variables and pose control is studied in detail in this paper.Because the traditional method is difficult to apply to the inverse kinematics of robot with redundant degree of freedom,the particle swarm optimization is used to transform the joint variables into the position coordinates of the particle,then the optimal particle is searched iteratively.On this basis,this paper designs two ways of trajectory planning for robot,and compares the advantages,disadvantages and the applicable scenes in detail,which is helpful to the realization of actions and the execution of tasks for robot in the simulation platform.By enabling robot to accomplish typical family tasks in the simulation environment,such as grasping the cup,opening the drawer and serving a drink for the user,we can test the effectiveness of each functional module and the ability of the simulation platform to simulate complex tasks,then verify the feasibility of using simulation platform to simulate robot performing daily tasks.In addition,we can analyze and evaluate the actions,control algorithms and service strategies of robot according to the simulation result,then make adjustments and improvements,thus we can obtain efficient task execution schemes and improve the service ability and level of robot. |
PDF Full Text Request