Design and optimisation of an underfloor energy harvesting system

Piezoelectric stack energy harvesters have the advantage of generating electrical power for small amounts of deflection. In an underfloor energy harvesting application, this is advantageous as the device does not disrupt the gait of pedestrians overhead. Low frequency, stochastic loading patterns are typical of underfloor applications, where the precise gait, weight and number of pedestrians are unpredictable. Energy harvesters targeting this type of directly applied load are faced with the challenge of providing sufficient force for power generation and determining appropriate matching of the piezoelectric stack to the circuit being powered, to achieve the highest possible energy conversion efficiency. The latter is particularly challenging in the case of stochastic input, due to the piezoelectric stack's dependence on frequency. To address these challenges, a force amplification frame is used in conjunction with a piezoelectric stack to provide an increase of up to 10 times of the applied load to the stack. A method of modelling the energy harvester is presented to optimise the device to suit a range of typical inputs, including walking, jogging and multiple pedestrian loading conditions. The methods used are capable of incorporating measured data from existing support platforms and the behaviour of interfacing power management circuitry into the optimisation loop, providing an accurate means of predicting the response of the energy harvesting system under typical operating conditions. The model is experimentally validated, and the results used to demonstrate a potential power increase of a factor of 21 when compared to an off-the-shelf piezoelectric stack. (C) 2018 Elsevier B.V. All rights reserved.