Research On The Key Technology On Of Multi-robit System Motion Planning

With the development of science and technology,robots are gradually becoming more intelligent and automated,and application scenarios are becoming more complex.For example,in the field of intelligent storage,the multi-robot motion planning technology,which realizes robot path planning and trajectory planning,has become an important research orientation in the process of robot development.In previous studies,kinematic constraints were not considered in path planning algorithms and the algorithms were not efficient when dealing with complex problems.Besides,the trajectory planning algorithms were difficult to extend to complex systems and tended to fall into local optimal solutions.To solve these problems,this paper proposes a multi-robot motion planning system based on path planning and trajectory optimization,which completes the motion planning of multiple robots through three parts: path planning,trajectory optimization,and iterative optimization.The system also takes into account kinematic constraints and planning capability for complex spaces,which can be widely used in fields such as intelligent storage.The three parts of the system correspond to two new path planning algorithms in Chapter 3 and a new trajectory optimization algorithm and a new iterative optimization algorithm within Chapter 4,respectively,as follows:1.To address the problem of low planning efficiency of path planning algorithms in complex environments,this paper proposes a graph search-based conflict classification and elimination algorithm optimization scheme,which adds corresponding constraint sets to each conflict type based on the characteristics and eliminates the current conflict and other path conflicts or predictable conflicts,so as to improve the algorithm planning efficiency.2.To address the problem of low planning efficiency of path planning algorithms in environments with specific demands,this paper proposes a graph search-based priority rule conflict resolution algorithm optimization scheme to dynamically adjust the robot priority according to the actual demand,so that the planned paths can be applied to specific environments.And this paper fuses the two novel algorithms as a discrete path planning algorithm in a motion planning system.Even in a large space and high-complexity environment,the algorithm can make the shortest discrete path for each robot for the motion planning system.3.To address the problem that path planning fails to solve dynamics constraints and trajectory planning cannot be extended to complex systems,this paper proposes a multi-robot trajectory optimization scheme based on support vector machines and Bezier curves.The algorithm converts the complex multi-robot trajectory planning problem into a single-robot optimal curve fitting problem in the safety space by dividing the safety space and Bezier curve fitting,and thus optimizes the discrete path of multi-robot into a curve trajectory satisfying the dynamics.4.To address the problems that the motion planning system maybe fall into the local optimal solution easily,this paper proposes a trajectory iterative optimization scheme based on a genetic algorithm.It further strengthens the planning ability of the motion planning system by applying the genetic algorithm to the Bezier curve fitting process and re-iterating the optimization results of the previous optimization.