This paper proposes a hybrid approach for switched reluctance motor (SRM) based water pumping system. The proposed hybrid method is combination of both the Northern Goshawk Optimization (NGO) and Finite Basis Physics-Informed Neural Networks (FBPINNs). Hence, it is named as NGO-FBPINNs. The NGO method is employed to better control among the three level boost converter (TLBC) and FBPINNs is predict the optimal control of the TLBC. In this system the water is pumped using a 4-phase SRM driven by a midpoint converter. An intermediary power conversion stage called a TLBC is positioned between the motor-pump and the solar PV array. The proposed method is very efficient because of the TLBC's small inductor size, wide operating voltage range, and low voltage stress across devices. The proposed method implemented in MATLAB platform and its efficiency is compared with various existing techniques like Heap-Based Optimizer (HBO), Wild horse optimizer (WHO) and Particle Swarm Optimization (PSO). The proposed approach NGO-FBPINNs obtains high efficiency than existing approaches. The proposed system's effectiveness is 95% it is higher efficiency than other system.

中文翻译：

---

采用混合方法的太阳能供电的基于开关磁阻电机的水泵系统

本文提出了一种基于开关磁阻电机（SRM）的水泵系统的混合方法。所提出的混合方法是北方苍鹰优化（NGO）和有限基础物理信息神经网络（FBPINN）的结合。因此，它被命名为 NGO-FBPINN。采用 NGO 方法更好地控制三电平升压转换器 (TLBC)，并且 FBPINN 预测 TLBC 的最优控制。在此系统中，使用由中点转换器驱动的 4 相 SRM 泵送水。称为 TLBC 的中间功率转换级位于电机泵和太阳能光伏阵列之间。由于 TLBC 的电感尺寸小、工作电压范围宽且器件间的电压应力低，因此所提出的方法非常有效。所提出的方法在 MATLAB 平台上实现，并将其效率与各种现有技术（如基于堆的优化器（HBO）、野马优化器（WHO）和粒子群优化器（PSO））进行了比较。所提出的方法 NGO-FBPINNs 比现有方法具有更高的效率。所提出的系统的有效性为95%，比其他系统的效率更高。