Genetic Algorithm Optimization for Mems Cantilevered Piezoelectric Energy Harvesters

A design optimization based on genetic algorithms for increasing power output performance of MEMS piezoelectric cantilevered energy harvesters has been performed. By employing genetic algorithms, the piezoelectric energy scavenging system with optimized shape and lateral geometries of a unimorph cantilever beam and a proof mass is simulated to achieve a power response improvement of ~1.8× in comparison with a reference system composed of a rectangular cantilevered structure. In particular, by optimizing the lateral geometries of a trapezoidal cantilever with proof mass, the power scavenged from a unit volume of the piezoelectric material can be enhanced to 162%. In addition, the output power per surface area from an optimized structure with a reverse trapezoidal cantilever enhances 184% based on theoretical analysis that includes damping effects.