Photovoltaic systems have a nonlinear characteristic in which there is one optimum operating point called Maximum Power Point (MPP). However, when PV panels are partially shaded by surrounding objects, there are several MPPs, of which one of them is Global MPP (GMPP). Therefore, conventional Maximum Power Point Tracking (MPPT) algorithms get trapped into local MPPs. As a result, a multitude of Global MPPT (GMPPT) algorithms have been proposed. An outstanding GMPPT algorithm as well as the fast-tracking speed should find GMPP in complicated shading patterns where not only there are lots of MPPs, but also the peaks are close together. Therefore, in this paper, a novel GMPPT based on firework algorithm is proposed which is able to find GMPP in complicated shading patterns with fast tracking speed. Moreover, the firework is combined with Perturb-and-Observe (P&O) algorithm to reduce the computational effort in a way that the firework is only used to recognize GMPP; afterwards, P&O algorithm completes the tracking. Furthermore, the variable sampling time technique, based on the system settling time, speeds up the tracking process considerably. Finally, the proposed method is compared with previous works, simulated, and implemented on an experimental setup to prove its superiority.

中文翻译：

---

基于烟花的具有可变采样时间的GMPPT用于光伏系统

光伏系统具有非线性特性，其中存在一个称为最大功率点（MPP）的最佳工作点。但是，当PV面板被周围的物体部分遮挡时，会有多个MPP，其中之一是Global MPP（GMPP）。因此，传统的最大功率点跟踪（MPPT）算法被困在本地MPP中。结果，提出了多种全局MPPT（GMPPT）算法。出色的GMPPT算法以及快速跟踪的速度应该可以在复杂的阴影模式下找到GMPP，在这种阴影模式下，不仅有很多MPP，而且峰之间也很靠近。因此，本文提出了一种基于烟火算法的新型GMPPT算法，该算法能够快速跟踪复杂阴影模式下的GMPP。此外，烟花与扰动和观察（P＆O）算法相结合，减少了计算量，从而使烟花仅用于识别GMPP。之后，P＆O算法完成跟踪。此外，基于系统建立时间的可变采样时间技术可大大加快跟踪过程。最后，将所提出的方法与以前的工作进行了比较，进行了仿真，并在实验装置上实施，以证明其优越性。