Research And Simulation Of Multi-robot Cooperation Based On Agent

Multi-robot cooperation is the key part to multi-robot technology research, and multi-robot hunting movement is a typical technology platform to study the technology of multi-robot cooperation. Multi-robot's cooperative hunting is becoming a hot issue and its technology is becoming an important research direction, with robots from the traditional "industrial robot" to "sensible robot" to "intelligent robot". Meanwhile, it is becoming the key to this issue that target robots'escaping strategy & hunting robots's hunting strategy, and how to achieve these more effectively.This paper studies multi-robot cooperation in hunting mobile targets. An analysis is made on the number of the robots, speed, initial conditions, hunting movement state and the conditions necessary for successful hunting. Dynamic adjustment is made to the escape strategy of target robot and capturing strategy of capturing robot. The result shows that the escape strategy featured by principle of maximum angle is put forward and the capturing strategy featured by a combination of head block and virtual potential point is also proposed. Furthermore, simulation of both strategies is carried out and the results show that the dynamic adjustment strategy proposed in the paper is effective.In order to distribute multi-robot system resources scientifically and reasonably, finish the task efficiently and in organization, this paper studies the agent alliance mechanism applied to multi-robot cooperation system task allocation problem to multi-robot system stalking assignment distribution problems, analysises from two aspects respectively of a single stalking task allocation and multi-task assigned, the speed, quantity, distance and task group members the relative position of the influence factors of assigned tasks are discussed, the mathematical model, genetic algorithm is used to optimize the solution, and simulation experiment is carried out, and the simulation results in a certain extent shows the multi-robots cooperative hunting task allocation model is effective.