Multi-robot Cooperative Movement Using Real-time Heuristic Algorithm

Mobile robots are widely used in various fields such as Unmanned Ariel Vehicle(UAV),Unmanned Ground Vehicle(UGV),and Unmanned Underwater Vehicle(UUV).Single-robot systems can only accomplish relatively simple tasks,and such system is hard to be applied to plenty of complex and large-scale scenarios,consequently,multi-robot systems emerge as requested.Cooperation among multiple robots can not only promote completion efficiency and enhance system reliability,but also be able to take over more complex tasks.As a result,multi-robot systems are widely applied into various fields such as automatic manufacturing system,search and rescue,and environmental monitoring,etc.For the multi-robot systems,robot path planning plays a key role in terms of whether tasks can be successfully accomplished.However,due to interferences in the actual environment,how to control robots and make path planning in real time,and how to avoid obstacles dynamically to make sure robots reach the target safely are urgent issues to be solved.This paper focuses on the use of formal methods for robot path planning.Temporal logic,such as Computation Tree Logic(CTL),Continuous Stochastic Logic(CSL),and Linear Temporal Logic(LTL)has been applied to express complex high-level planning specifications.However,traditional methods require the input of global environment information in advance and cannot effectively avoid the dynamic obstacles added in real time.Also,the centralized path planning method requires central controller to plan the path,which lacks in computational efficiency and robustness.To overcome abovementioned shortcomings,this paper proposes a real-time online path planning and control method for multi-robot systems.This method utilizes Computational Tree Logic(CTL)to define the motion specification and grid-based map to make representation.Robots have the ability of communication and sensing.On the premise of unknown global environmental information and without a central controller,in order to reach the target position at the least cost,and avoids dynamic obstacles,robots adjust the route in real time according to environmental changes.Moreover,Manhattan method is used to score the path of each robot.Specifically,costs of the robot from the current position to the next are constant,and robots estimate the costs from the next position to the target position.For single-robot systems,under the same static environment,paths of this method are the sameas those found by offline heuristic algorithm.The effectiveness of this method is verified by programming and simulation.For multi-robot systems,this paper focuses on the path planning of cooperative robots to accomplish aggregation tasks.Taking the multi-robot systems composed of two robots and four robots as examples,adopting distributed control structure,cargo handling is simulated by programming simulation to verify the correctness of real-time cooperative control algorithm.Performance and effectiveness of the method are verified by adding dynamic obstacles dynamically during the motion process.