Effect of nonlinearities and objective function in optimization of an energy harvesting device

This work presents a study of the impact of the linearity assumption of the mechanical model in the overall performance of an energy harvesting piezoelectric beam. Also, a brief assessment of geometrical optimization solutions using different objective functions is presented. The mechanical model of the harvester is based on both linear and nonlinear variants of the electrical and mechanical constitutive equations for the piezoelectric material. The nonlinear elastic, damping and electromechanical coupling parameters are obtained via least squares identification using physical experimentation; the experimental tests are performed at different ground excitation accelerations. The computational optimization of the harvester is done using the genetic algorithm implemented in Matlab. Different objective functions are tested, i.e. broadband maximum peak power, maximum power at a particular frequency and broadband mean power; the influence of the selection of each of them in the total recovery of the power of the device is analyzed. The most suitable function to recover the vibratory energy from conventional transport vehicles is found.