For PV in case of complex application environment has the highly unlinear, several local peak power and measured noise characteristics. Artificial fish school algorithm(AFSA) has the ability of overcoming local peak power and obtaining global best, very suitable for solving this optimization problem of non-linear extreme value function. Analyzed the PV output characteristics under complex application environments,proposed the AFSA maximum power point tracking (MPPT) control algorithm based on least square filtering to overcome the effects of measured noise. Simulation results indicate that the proposed MPPT control method can suppress the influence of the measurement noise effectively, and can improve the array output efficiency.