Path planning of robot using genetic annealing algorithm

Abstract

The paper presents a new approach, which uses genetic annealing algorithm to solve multi-robot path planning in dynamic environment. The algorithm avoids the common defect of early convergence in genetic algorithm, and can converge faster than simulated annealing algorithm or genetic algorithm alone. We have designed an effective path planning crossover operator pointing to the characteristics of complex environment. Simulation and real experiment results show that the method is effective in multi-robot path planning.