Modeling and Optimization of Piezoelectric Energy Harvesting System Under Dynamic Loading

In this work, energy is harvested from circular piezoelectric materials with different boundary conditions under dynamic loading. The output of the proposed system is investigated by connecting piezo with full wave bridge rectifier (FWBR) and voltage doubler (VD) circuits. The proposed setup can be used to apply dynamic load with varying magnitude. The boundaries of the circular piezoelectric material are 3D printed using Acrylonitrile butadiene styrene (ABS) material with FDM 3D printing machine to fabricate its simple support (SS), fixed support (FS) and circular support (CS). The proposed setup utilizes a stepper motor, a display unit, a load sensor, eccentric mechanism to convert rotary motion into reciprocating motion, power supply and a compatible circuit on the printed circuit board. It is observed experimentally that the energy harvested is maximum, in case of circular support among the three applied boundary conditions and validated mathematically. The output response of the circular support boundary condition is analyzed using various combinations of three input parameters, i.e., varying dynamic load, number of strokes, two circuits and a total number of 120 experiments were performed. Artificial neural network has been applied for modeling and training between input three parameters and one output response, i.e., energy harvested. Further, genetic algorithm is applied on the ANN model to maximize the output of the system w.r.t input parameters and experimentally validated. As number of strokes (26) at dynamic load of 44 N with circular supported piezo patch connected with voltage doubler circuit gives the maximum output of 16.09 V further, when the system is tested on these parameters experimentally, it gives the 15.26 V.