Neural Network-Based Finite-Time Control for Solar Tracking Systems

Solar tracking systems improve the efficiency of photovoltaic (PV) and concentrated solar power technologies by maximizing solar energy capture. However, their performance is affected by uncertainties such as environmental disturbances, mechanical imperfections, and nonlinear system dynamics. This paper proposes a Neural Network-Based Robust Terminal Sliding Mode Control (NN-TSMC) approach to enhance tracking accuracy and robustness. The TSMC ensures finite-time convergence and disturbance rejection, while an adaptive NN compensates for unknown system uncertainties in real time. A Lyapunov-based approach is proposed to ensure the controlled system's stability. Simulation results demonstrate that the NN-TSMC outperforms conventional controllers in tracking precision, reducing chattering effects, and adapting to uncertainties. The proposed method provides a reliable, intelligent control solution for solar tracking applications, making it suitable for real-world deployment in dynamic environments.