Inclusion of bypass diodes at the output terminal of the PV array mitigates the effect of partial shading (PS) but causes multiple peaks of power at the output. The conventional hill climbing and perturb and observe algorithms cannot track the optimal point during partial shading phenomena for multiple peaks corresponding to the different shading pattern on the Power-Voltage (P-V) curve. Fuzzy logic controller and artificial neural network-based methods for Maximum Power Point Tracking (MPPT) provide satisfactory results but at the cost of increased memory and computational burden. Recent work to incorporate exploration and exploitation phenomena of nature-inspired algorithms to track optimal power point have shown encouraging results by preventing convergence to local maxima and posing less burden on the processor. However, due to performance variation between different algorithms of this category newer algorithms with improved performances are still a requirement. In this paper, a novel most valuable player algorithm (MVPA) has been used to track the optimal operation point for extracting maximum power from a solar PV system. The algorithm's performance is compared with the commonly employed particle swarm optimization (PSO) and the recently proposed Jaya algorithm's modified form. It is observed that the proposed algorithm outperformed both the algorithms with a considerable improvement in terms of tracking speed, power tracking efficiency, robustness, faster decision for convergence after tracking the maximum power and lesser number of power fluctuations for different shading patterns.

中文翻译：

---

基于最有价值玩家算法的部分遮蔽光伏发电系统的最大功率点跟踪

在 PV 阵列的输出端子包含旁路二极管可减轻部分遮蔽 (PS) 的影响，但会导致输出功率出现多个峰值。传统的爬山和扰动观察算法无法跟踪对应于功率电压 (PV) 曲线上不同阴影模式的多个峰值的局部阴影现象期间的最佳点。用于最大功率点跟踪 (MPPT) 的模糊逻辑控制器和基于人工神经网络的方法提供了令人满意的结果，但代价是增加了内存和计算负担。最近结合自然启发算法的探索和利用现象来跟踪最佳功率点的工作通过防止收敛到局部最大值并减少处理器负担而显示出令人鼓舞的结果。然而，由于该类别的不同算法之间的性能差异，仍然需要具有改进性能的较新算法。在本文中，一种新颖的最有价值播放器算法 (MVPA) 已被用于跟踪从太阳能光伏系统中提取最大功率的最佳操作点。该算法的性能与常用的粒子群优化 (PSO) 和最近提出的 Jaya 算法的修改形式进行了比较。可以看出，所提出的算法在跟踪速度、功率跟踪效率、鲁棒性、跟踪最大功率后收敛决策更快以及不同着色模式的功率波动次数更少方面均优于这两种算法。