Dual-Axis Solar Tracking Technique Combining MPC, PID, and ANN Control with Dynamic Controller Selection

Accurate sunlight tracking is essential for maximizing photovoltaic power generation. However, conventional tracking systems based on proportional-integral-derivative (PID) and model predictive control (MPC) face significant challenges. While PID controllers are simple and reliable, they often lack accuracy in the presence of disturbances and nonlinearities. In contrast, MPC controllers offer higher precision but come with increased computational complexity. This paper proposes a novel dual-axis solar tracking system that integrates PID and MPC with an artificial neural network (ANN) to address these limitations. The ANN dynamically selects the most suitable controller in real-time based on environmental conditions and system performance. Additionally, a new dual-axis mechanical structure is introduced to further enhance tracking accuracy. Simulation results demonstrate that the proposed hybrid system outperforms traditional tracking methods that employ PID and MPC in parallel, delivering improved responsiveness and robustness under varying climatic conditions. The system achieves minimal tracking error, with low average absolute errors of ۰.۰۱۶۱° in altitude and ۰.۰۱۹۹° in azimuth. These findings validate the effectiveness of the MPC-PID switching strategy in achieving high-precision solar tracking, positioning it as a promising solution for advanced photovoltaic applications.

کلیدواژه ها:

Artificial Neural Networks ، Optimal control ، Solar tracking system ، photovoltaic systems ، MPC Controller