阴影条件(partial shading condition, PSC)下光伏系统的而功率-电压(P-V)特性曲线呈多峰形，易造成常规最大功率跟踪(maximum power point tracking, MPPT)算法陷入局部最大功率点(local maximum power point, LMPP)的问题。本文采用一种新颖的启发式算法，即交互式教-学优化算法(interactive teaching-learning optimization, ITLO)来实现光伏系统PSC下的MPPT。ITLO在原始教-学优化算法(teaching-learning based optimization, TLBO)的基础上，采用多个班级同时进行教与学，以实现更广泛的全局搜索，提高最优解的质量；同时，所有班级的教师与教师、学生与学生间引入小世界网络(small world network, SWN)机制进行交互学习，以实现更深入的局部探索，有效避免了算法陷入LMPP，并提高其收敛速度和收敛稳定性。文中进行了恒温变光照强度和变温变光照强度两个算例的研究。仿真结果表明，与增量电导法(incremental conductance, INC)和遗传算法(genetic algorithm, GA)相比，ITLO能在PSC下最快速，最稳定地获取最大光能。
英文摘要:
Photovoltaic systems are often affected by partial shading condition(PSC) which result in its power-voltage (P-V) characteristics present multiple peaks. Under this condition, the conventional maximum power tracking (MPPT) techniques are easy to fall into a local maximum power point (LMPP). This paper designs a novel MPPT algorithm called interactive teaching-learning optimization (ILTO). Based on the original teaching-learning based optimization (TLBO), ILTO employs the multiple classes to carry out the teaching-learning which aims to achieve a wider exploration and to improve the quality of the optimal solution. Meanwhile, the small world network (SWN) is introduced for interactive learning among teachers or students from different classes to realize a deeper exploitation, which presents the algorithm from trapping at LMPP as well as enhances the convergence speed and convergence stability of ILTO. In this paper, two cases studies are implemented, e.g., (a) constant temperature and varying solar irradiation; and (b) varying temperature and varying solar irradiation. Simulation results demonstrate that, compare with incremental conductance (INC) and genetic algorithm (GA), ILTO could achieve the fastest and most stable global MPPT under PSC.