For the maximum utilization of solar energy, photovoltaic (PV) power systems are operated at the maximum power point (MPPT) under changing environmental conditions, and maximum power point is generally achieved using traditional methods. However, the occurrence of multi-peak due to partially shaded conditions increase complexity of the tracking prosses and conventional technique can't guarantee convergence to global maximum power point always. This paper proposed a MPPT method for solar energy system using a novel flower pollination (FP) algorithm. This paper reports the conception of FP algorithm by emphasizing the importance of the Levy flight in affecting the proposed method's convergence. In order to verify the effectiveness of the proposed method, the MPPT control model of the PV system is established in MATLAB and compared with two well established techniques. The simulation results indicate that the proposed MPPT technique can very quickly track the varies of external environment and effectively handle the partially shaded condition.