No abstract

A modified genetic algorithm for global path planning of searching robot in mine disasters is proposed in this paper. The grid model is built based on the mine GIS (Geographic Information System) which can be obtained from the mine in advance. Using the position information feedback from the Ant Colony Optimization and priority grouping, we present a new efficient method to generate the initial population. Four traditional genetic operators and a fitness function are designed to find the optimum path planning. To avoid the premature, we make some changes to the mutation operators, and self-adaptively adjust the probabilities of crossover and mutation. The simulation is carried out in MATLAB, and the result verifies that it can acquire better collision-free path in higher speed.Index Terms -searching robot, global path planning, genetic algorithm, position information feedback and priority grouping, self-adaptive probabilities