Hybrid Learning Approach based on Multi-Objective Behavior Coordination for Multiple Robots

Abstract

The paper researches the collision avoidance and target tracing problem for multi robots in a dynamic environment. Robot's motion is controlled by the multi-objective behavior coordination based fuzzy inference rules. In order to obtain local and global optimal behaviors, a hybrid learning approach is further proposed. Each fuzzy rule is expended to have multiple possible strategies. The selection probability of strategies is updated by the Learning Automaton, and output parameters of fuzzy rules are updated by the Steady-state Genetic Algorithm. Simulations are done to verify the proposed approach, and simulation results prove the feasibility of the proposed approach.