It is essential to change the load characteristic under different light and temperature conditions in order to track the maximum power point of solar cell rapidly and accurately, according to the non-linear characteristics of photovoltaic cells. For general genetic algorithm in the process of maximum power point tracking MPPT, there are disadvantages such as premature and slow convergence insufficiency, in this paper, a new quantum genetic algorithm combined with perturbation method, based on the Buck-Boost circuit, is used on the circuit topology of alternatives instead of traditional buck circuit or boost circuit, and this method is established and simulated under Matlab/Simulink. Experimental results show that the algorithm has better advantage searching and tracking ability, good control precision, and effectively restrain the fluctuation near the maximum power point, to prove the accuracy and effectiveness of the proposed algorithm.