This study proposes a hybrid global maximum power point tracking (GMPPT) scheme integrating an extreme learning machine with 0.8V_oc technique for PV system. An attempt is made to employ an anti-windup fractional-order proportional-integral controller for the MPPT. The controller parameters were tuned using an enhanced salp swarm algorithm. The algorithm integrates via an accurate detection scheme that distinguishes partial shading conditions (PSCs) from an irradiance uniform change. Furthermore, the computed irradiance is used to update PV array open-circuit voltage (V_{oc_Array}), preventing temperature and irradiance sensors from being used. Its performance was studied compared with MPPT controllers, i.e., deterministic particle swarm optimization, hybrid PSO, and Lagrange interpolation PSO. The proposed MPPT technique proved its ability to track GMPP with an average tracking efficiency of 99.20% and 99.10% for uniform and PSCs, respectively. The proposed scheme has significant speed and accuracy in tracking GMPP for complex PSCs and uncertain weather conditions. Irrespective of the environmental uncertainties, it has an average voltage tracking percentage error within ± 1% for ten hours test profile. The proposed technique is explored on OPAL-RT 4510 platform. The results depict its ability in GMPP tracking with an average tracking efficiency and tracking time of 99.15% and 0.12 s, respectively.

本研究提出了一种混合全局最大功率点跟踪 (GMPPT) 方案，该方案将极限学习机与 0.8 V _oc技术集成到光伏系统中。尝试为 MPPT 使用抗饱和分数阶比例积分控制器。使用增强的 salp swarm 算法调整控制器参数。该算法通过准确的检测方案进行集成，该方案将部分阴影条件 (PSC) 与辐照度均匀变化区分开来。此外，计算出的辐照度用于更新光伏阵列开路电压 ( V _{oc_Array})，防止使用温度和辐照度传感器。将其性能与 MPPT 控制器（即确定性粒子群优化、混合 PSO 和拉格朗日插值 PSO）进行了比较。所提出的 MPPT 技术证明了其跟踪 GMPP 的能力，均匀和 PSC 的平均跟踪效率分别为 99.20% 和 99.10%。所提出的方案在跟踪复杂 PSC 和不确定天气条件的 GMPP 方面具有显着的速度和准确性。不考虑环境的不确定性，十小时测试曲线的平均电压跟踪百分比误差在 ± 1% 以内。所提出的技术在 OPAL-RT 4510 平台上进行了探索。结果描述了其 GMPP 跟踪能力，平均跟踪效率和跟踪时间分别为 99.15% 和 0.12 s。