Research On Multi-robot System Formation

Research on multi-robot system formation issue includes two parts:formation generating and formation control. Formation generating is the method to transform a multi-robot system with an unordered initial state into a regularly needed formation. Formation control is the strategy that a multi-robot formation moving towards the target area while keeping the formation shape and accommodating the current working environment constraints (such as obstacles or spaces physical restrictions). Study of multi-robot formation has great significance. Multi mobile robots in specific formations can be applied to industry, aerospace, military, etc.In the multi-robot formations, the numbers of the robots will be decades to hundreds so that they are large scale robot systems. In order to reduce the calculated amount of generating formation, a classification-based search method is provided to solve arbitrary formation generating problems for large-scale robot teams. The workspace will be divided into square cells and the robots will be classified into3different categories. Robots in different categories have different assignment priorities. It can avoid global scope search and reduce the computation complexity of the assignment process.The main features of the large scale robot systems are that lower performance of a single robot, higher request to the formation control, higher request to the efficiency of the algorithm. Therefore, distributed robot systems are adopted. More suitable behavior-based control method is put forward for large-scale robot systems. Avoiding obstacles and keeping formation are implemented while robot teams moving to the targets.The validity of all mentioned above has been tested and verified by simulation experiments.