The P-U curves of partially shaded photovoltaic arrays are of multi-peak

which makes it necessary to design a method for photovoltaic multi-peak maximum power point tracking(MPPT)to realize the maximum power output and to improve the efficiency of photovoltaic power generation. Compared with the particle swarm optimization(PSO)algorithm

the quantum-behaved particle swarm optimization(QPSO)algorithm has advantages such as a faster convergence rate and global convergence. An improved method for photovoltaic multi-peak MPPT based on the QPSO algorithm is proposed. The method applies the QPSO algorithm to the global search for the maximum power point. The total number and the voltages of the particles are initialized according to the distribution character of the extreme points on the P-U curves of the partially shaded photovoltaic arrays. A convergence criterion which is more suitable for photovoltaic multi-peak MPPT is proposed according to the characteristics of the particles' personal best voltages when the QPSO algorithm converges. The simulation test shows that the improved method can realize multi-peak MPPT quickly and effectively

and avoids the non-convergence problem. It also has the ability to respond to the change of the light condition

and improves the efficiency of the photovoltaic power generation when partially shaded.