The problem of path panning in mapped environment is well-known in roboties. Given an object with an initial pasition and a goal position, and a set of obstacles located in space, the problem is find a optimal or feasible path for the object to move from the initial position to the goal position which avoids colliding with obstacles along the way. This paper proposes a new strategy using the Genetic algorithm (GA) to solve the collision-free path planning problem. Genetic algorithm is a searching and optimizing algorithm based on the mechanics of natural genetics and natural selection. The proposed algorithm introduces feature map describing the attributes of the selected path strings and path recombination operator having the effect of crossover and mutation, and demonstrates the effectiveness and efficiency through a simulation to reach a grobal optimal or feasible path solution.
