Research And Application Of Motion Planning For Robots Under Temporal Task Constraints

Path planning is one of the key technologies in the research of robots.The study of path planning technology can promote the practical application of robots in life.Most traditional path planning methods focus on the mission planning between two points,which cannot satisfy the requirements of temporal task.At present,the linear temporal logic(LTL)path planning methods can complete complex multi-point mission planning,but they are suitable for the static environment only,not the complex environment that changes frequently.At the same time,the current path planning methods are to search path based on the discrete environment model and do not implement the motion control of transferring discrete paths to continuous trajectory.Therefore,the robot motion planning method that satisfies temporal task requirement in complex environments is very important.The LTL theory provides theoretical support for the description of the robot temporal tasks.The LTL formula can describe the complex temporal task according to its syntax and semantics.In the paper,the motion planning method under temporal task constraints in complex environment is studied.Since it is difficult to construct a network topology with limited node states for an outdoor complex environment,the LBS-based motion planning under temporal task constraints is firstly studied to solve this problem.LBS is a location-based service that combines technologies such as geographic information and mobile communication.According to the nodes of temporal tasks,a finite-state transition system is built based on Baidu map LBS,temporal tasks are described by LTL formula,and Product automaton containing environment information and task information is generated.We propose a path search algorithm based on Dijkstra algorithm to search for the optimal discrete path on the Product automaton,and transfer the discrete path to continuous trajectory according to the Baidu map LBS.Then,according to the indoor dynamic environment and the robot dynamic model,a temporal task motion planning method based on the robot dynamic model is studied.The indoor work environment of robots based on cell decomposition method is constructed as an environment model and a weighted transition system,and a global path search algorithm based on A* algorithm is proposed to search for a discrete path,which improves search efficiency.In order to transfer the discrete path into continuous trajectory,a robot state transition control method that incorporates obstacle avoidance strategies is designed on the basis of a robot dynamic model.Finally,based on SLAM,a robot motion planning system under temporal task constraints was designed to implement the temporal task motion planning method satisfying the robot dynamic model.The system uses Monte Carlo Localization algorithm to complete the robot self-localization.At the same time,the computer operation interface of setting up temporal tasks and double obstacle avoidance mode are designed to achieve the easy operability and applicability of the system.The above motion planning for robot method under temporal task constraints is simulated and experimented in this paper.The results show that the method proposed in this paper can complete the temporal task motion planning in complex environments,realize the hybrid control of the integration of the high-level task planning and the low-level motion control,and promote the application of the robot in the complex environment.