Scholarlink

<p>The increasing deployment of standalone photovoltaic (PV) in rural and off-grid regions is crucial for addressing sustainable water supply and energy access challenges to remote areas. However, these systems demand high-performance Maximum Power Point Tracking (MPPT) controllers capable of efficiently harnessing solar energy under fluctuating environmental conditions. Conventional MPPT techniques, such as perturb and observe (P&amp;O) and incremental conductance (INC), often suffer from slow response times, steady-state oscillations, and reduced efficiency, limiting their suitability for real-world deployment. This paper represents a novel hybrid MPPT controller that integrates the predictive capability of artificial neural networks (ANNs) with the robustness and finite-time convergence of finite-time sliding mode control (FTSMC). Unlike existing approaches, the proposed ANN–FTSMC achieves fast convergence within a settling time of 0.2&#xa0;s, with 99.97% efficiency, and minimal current oscillations of 0.01%, as demonstrated in MATLAB/Simulink simulation results. Comparative results show that the proposed method outperforms conventional MPPT algorithms by up to 15% in dynamic efficiency and significantly reduces steady-state ripples. Real-time implementation on an NXP embedded platform validated the practical feasibility and scalability of the proposed approach with performance consistent with the simulations. These outcomes position the ANN-FTSMC as a robust, and deployable solution for intelligent MPPT in standalone PV water pumping systems (PVWPS).</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Novel Hybrid Artificial Neural Network and Finite-Time Sliding Mode Control Strategy for MPPT in PV Water Pumping Systems with Real-Time Implementation

  • 期刊论文

  • 发表日期：

    2025年12月2日

查看全文
  • Salma El Aamery,
  • Rachid El Bachtiri,
  • Rafika El Idrissi

摘要

The increasing deployment of standalone photovoltaic (PV) in rural and off-grid regions is crucial for addressing sustainable water supply and energy access challenges to remote areas. However, these systems demand high-performance Maximum Power Point Tracking (MPPT) controllers capable of efficiently harnessing solar energy under fluctuating environmental conditions. Conventional MPPT techniques, such as perturb and observe (P&O) and incremental conductance (INC), often suffer from slow response times, steady-state oscillations, and reduced efficiency, limiting their suitability for real-world deployment. This paper represents a novel hybrid MPPT controller that integrates the predictive capability of artificial neural networks (ANNs) with the robustness and finite-time convergence of finite-time sliding mode control (FTSMC). Unlike existing approaches, the proposed ANN–FTSMC achieves fast convergence within a settling time of 0.2 s, with 99.97% efficiency, and minimal current oscillations of 0.01%, as demonstrated in MATLAB/Simulink simulation results. Comparative results show that the proposed method outperforms conventional MPPT algorithms by up to 15% in dynamic efficiency and significantly reduces steady-state ripples. Real-time implementation on an NXP embedded platform validated the practical feasibility and scalability of the proposed approach with performance consistent with the simulations. These outcomes position the ANN-FTSMC as a robust, and deployable solution for intelligent MPPT in standalone PV water pumping systems (PVWPS).