Path Planning For Heterogeneous Robot System With Road Network Constraints

In the study of multi-robots,heterogeneous robot systems have been widely studied due to their individual performance characteristics that can complement each other.This thesis studies the path planning of heterogeneous robot systems.The main work is as follows:Firstly,based on a large amount of literature,the current research status of path planning for heterogeneous robot systems is analyzed.At present,the research on path planning of heterogeneous robot systems can be divided into two categories from the perspective of path planning of ground mobile robots: one is to allow ground mobile robots to move at any position on the ground,and to plan paths for ground mobile robots and drones at the same time;the other is to limit the range of motion of the ground mobile robot,but given its route artificially,only the route of the drone is planned.This thesis considers the limitations of the ground road network in practical applications and plans paths for ground mobile robots and drones.Then,the path planning problem in this thesis is modeled,and the implementation of the "two-step" strategy is introduced in detail.Due to the strong coupling between the routes' of mobile robot and drone,and the complexity of the ground road network,it is difficult to express the route of mobile robot using traditional constraints,that is,the coupling between mobile robot and drones is difficult to solve by constraints.Therefore,this thesis uses ant colony algorithm and genetic algorithm to solve the path of ground mobile robot and drone respectively.Both the path of the drone and the ground car are decoupled,and the two routes are affected by the combination of ant colony algorithm and genetic algorithm.Lastly,in order to further shorten the total flight distance of the drone,the drone is allowed to visit multiple target points at one time during the endurance time,and the model and algorithm of the path planning of the drone are extended.The effectiveness of the model and algorithm is verified through simulation.