Real-Time Optimization of Piezoelectric Energy Harvesting Using ANN-Based Maximum Power Point Tracking

Piezoelectric Energy Harvesting (PEH) offers a promising solution for powering low-power electronic systems exposed to ambient vibration. However, the fluctuating and nonlinear nature of mechanical input leads to continuous variations in electrical output, making Maximum Power Point Tracking (MPPT) essential for efficient energy extraction. Traditional fixed-duty or pulse-based control methods struggle to maintain MPPT under dynamic vibration conditions, resulting in significant power loss. This study proposes an Artificial Neural Network (ANN)-based MPPT controller capable of adaptively determining the optimal duty cycle for a DC-DC buck converter in real time. A comprising a piezoelectric bender, full-wave rectifier, and lithium-ion battery-was modeled and simulated in performed under varying vibration amplitudes and frequencies to generate training data for the ANN. The trained ANN achieved a high correlation coefficient (R = 0.99332), confirming its accuracy and generalization capability. Simulation results show that the ANN controller significantly stabilizes the rectifier voltage, enhances impedance matching, and improves battery charging performance. These findings demonstrate that ANN-based MPPT provides an efficient and robust solution for real-time power optimization in piezoelectric vibration energy harvesting systems.