Harrison Ambe, Alombah Njimboh Henry, de Dieu Nguimfack Ndongmo Jean
Department of Electrical and Electronics Engineering, College of Technology (COT), University of Buea, P.O.Box Buea 63, Cameroon.
Department of Electrical and Electronics Engineering, College of Technology, University of Bamenda, P.O. Box 39, Bambili, Cameroon.
Heliyon. 2023 Jul 20;9(8):e18434. doi: 10.1016/j.heliyon.2023.e18434. eCollection 2023 Aug.
The efficient operation of PV systems relies heavily on maximum power point tracking (MPPT). Additionally, such systems demonstrate complex behavior under partial shading conditions (PSC), with the presence of multiple maximum power points (MPP). Among the existing MPPT algorithms, the conventional perturb and observe, and incremental conductance stand out for their high simplicity. However, they are specialized in single MPP problems. Thus, due to the existence of multiple MPPs under PSC, they fail to track the global MPP. Compared with the conventional schemes, the modified conventional algorithms, and several existing MPPT variants introduce a trade-off between complexity and performance. To enhance the simplicity of the PV system, it is crucial to adapt the operation of the simple conventional algorithm to scenarios under PSC. To achieve such an adaptation, the power-voltage curve that conventionally admits multiple MPPs under PSC must be converted to an equivalent curve having only a single MPP. To address such a requirement, this paper introduces a novel approach to the fast determination of the MPP. A consistent methodology for reducing the complex multiple MPP problem of PV systems under PSC, to a single MPP objective, is put forward. Thus such reduction enhances the tracking environment for simple conventional MPPT algorithms under partial shading. Studies of the PV array behavior for 735 partial shading patterns revealed an interesting possibility of reducing the classical PV curve to 8.2620% of its actual area. The newly established area is an optimum power region that accommodates a single MPP. To arrive at such a reduction, an intelligent neural network-based predictor, incorporating a cost-effective and reliable solar irradiance estimator is put forward. Unlike existing methods, the approach is free from the direct and expensive measurement of solar irradiance. The predictor relies on the PV array current and voltage only to precisely determine the optimum power region of the PV system.
光伏系统的高效运行在很大程度上依赖于最大功率点跟踪（MPPT）。此外，此类系统在部分阴影条件（PSC）下表现出复杂的行为，存在多个最大功率点（MPP）。在现有的MPPT算法中，传统的扰动观察法和增量电导法因其高度简单而脱颖而出。然而，它们专门用于解决单最大功率点问题。因此，由于在部分阴影条件下存在多个最大功率点，它们无法跟踪全局最大功率点。与传统方案相比，改进的传统算法以及几种现有的MPPT变体在复杂性和性能之间进行了权衡。为了提高光伏系统的简单性，使简单的传统算法能够适应部分阴影条件下的场景至关重要。为了实现这种适应，必须将在部分阴影条件下通常允许多个最大功率点的功率 - 电压曲线转换为仅具有单个最大功率点的等效曲线。为满足这一要求，本文引入了一种快速确定最大功率点的新方法。提出了一种将光伏系统在部分阴影条件下复杂的多个最大功率点问题简化为单个最大功率点目标的一致方法。因此，这种简化增强了简单传统MPPT算法在部分阴影下的跟踪环境。对735种部分阴影模式下光伏阵列行为的研究揭示了将经典光伏曲线面积缩小至其实际面积8.2620%的有趣可能性。新确定的区域是一个容纳单个最大功率点的最佳功率区域。为了实现这种缩小，提出了一种基于智能神经网络的预测器，该预测器集成了具有成本效益且可靠的太阳辐照度估计器。与现有方法不同，该方法无需直接且昂贵地测量太阳辐照度。该预测器仅依靠光伏阵列的电流和电压就能精确确定光伏系统的最佳功率区域。
