The research of finding optimal observation place for inversing concrete gravity dam parameters was studied firstly through analytical sensitivity analysis theory, the conclusion indicated that the displacement of dam crest is the most suitable data for inversing dam parameters. Then a novel stochastic search optimization algorithm—Artificial Fish Swarm Algorithm (AFSA ) was presented, according to the incremental displacement of dam crest, it was applied to inverse the parameters of gravity dam and bedrock,at the meantime, the numerical stability of AFSA was also analyzed. Results of computational example show that the proposed inverse method can rapidly converge to the global optimum solution, and has the characteristics of global convergence and the strong ability of noise resistance. So it can be concluded that the inverse method of dam parameters based on AFSA is workable and reliable.
