Abstract:Based on the theory of the physical cellular automata (PCA) and genetic algorithm,an evolving cellular automata is put forward. When rock failure is simulated with PCA,the energy dissipative rate has to be chosen artificially. This model makes use of genetic algorithm to search for the fittest energy dissipative rate with the stress-strain curve. And it can simulate the homogeneity and the anisotropy of the rock conveniently. So if a stress-strain curve is given,its corresponding energy dissipative rate can be attained with this model and it demonstrates that the rock will show more plasticity with the increase of the energy dissipative rate.