Optimum network configuration design of a multi-beam vortex-induced vibration piezoelectric energy harvester

Abstract

The focus of this study is on multi-beam piezoelectric energy harvester (MPEH) network optimization with respect to its configuration. Indeed, the harvested power in the single piezoelectric energy harvesters (SPEHs) may not meet the power level necessary to many applications. In case of constraint in the beam dimensions, the issue of sufficient power generation can be undertaken by arranging a definite number of those harvester units as an MPEH network. The versatility of network configurations presented in the literature, consisting of the harvester beams and their series/parallel interconnections, are limited to only two scenarios including all-series and all-parallel connections of the beams. As will be shown, these classic configurations are not necessarily the optimum ones for a given vortex shedding (VS) frequency. In fact, the harvested power as a function of VS frequency can be maximized by adopting a special network configuration, referred to as the optimum network configuration (ONC). In this study, the general form of the harvested power for each VS frequency is presented in a closed-form expression as a function of the network parameters. An optimization problem is formulated to maximize the harvested power and to obtain the ONC, which is solved through a proposed algorithm named the network configuration optimizer (NETCOOP). In the proposed MPEH structure, electrical switch blocks are used to implement the ONC and to reconfigure the network according to each VS frequency. The obtained VS frequency-dependent ONC enables the designers to develop the best configuration at each fluid flow velocity. To highlight the advantages of this proposed algorithm, the performance of the ONC is compared with other configurations like the all-series and the all-parallel configurations. In the simulation results, the optimum harvested power is obtained for different VS frequencies and load resistances by applying the brute-force approach. It is revealed that the harvested power obtained by the NTECOOP algorithm matches the optimum performance.