Abstract: This paper presents an artificial fish-school algorithm based on adaptive Cauchy mutation.By adding best fish so far and Cauchy mutation operator to artificial fish school algorithm (AFSA) in evolution process, the ability of AFSA to break away from artificial fish stochastic moving without a definite purpose or heavy getting together round the local optimum solution is greatly improved.The proposed algorithm can greatly improve the ability of seeking the global excellent result, convergence property and accuracy.Several computer simulation results and theories show that the algorithm is significantly superior to basic AFSA.