Optimization of grid wall walking by genetic algorithm

Grid wall walking promises attractive 3-D mobility for multi-legged robots. The robot hangs the grids by its multi-legs so that the stable positioning is achieved. The grid environments are implemented by assembling iron wires in a matrix. The artificial environment can be good for the robots by considering robot´s ability. The grid wall will be attached on various wall and ceiling. The robot has to hang them in highly stable posture since it must not fall down. Multi-legged robots have various hanging styles and it is difficult to find the available and optimal posture. The nonlinear search problem of the optimal posture can be solved by GA(Genetic Algorithm). The optimal hanging postures are found out on all of inclined grid wall, and these postures can be switched from each other to walk the complicated wall terrain. This paper shows the method of optimization for grid wall walking and the results show the feasibility of our proposed 3-D environmental mobility.