Model predictive control for optimized piezoelectric energy harvesting under multimodal vibration excitation: theory and simulation

Abstract In this paper, a novel switching control strategy for multimodal piezoelectric vibration energy harvesting (PVEH) systems using model predictive control (MPC) is presented. Although conventional switching methods have been shown to be effective for single-degree-of-freedom systems, they are often unable to achieve optimal performance in multimodal vibration environments owing to the varying displacement amplitudes across different modes. To overcome this limitation, a mathematical model for multi-degree-of-freedom PVEH systems has been developed, and an MPC-based switching strategy has been proposed that optimizes energy-harvesting efficiency by predicting the system behavior over a finite time horizon. The proposed approach enables dynamic adaptation of the switching timing for both charge inversion and extraction switching based on the predicted future states. Numerical simulations demonstrate that the predictive switching method achieves a significantly higher energy-harvesting efficiency than the conventional method, with improvements observed in different scenarios with piezoelectric factors. The robustness of the method is validated by a parameter sensitivity analysis. A computational feasibility analysis also shows that the proposed strategy maintains real-time performance with control horizons of up to 35 units. These results suggest that the developed predictive switching control strategy offers a promising solution to improve energy-harvesting efficiency in practical multimodal PVEH applications.