Multi-objective optimisation methodology for self-sensing piezoelectric monomorph benders

Abstract

Charge-based self-sensing actuation (SSA) with piezoelectric actuators has achieved great strides of progress since their inception in 1992, with multiple proposals improving the estimation method through the use of efficient non-linearity and error compensations. The aim of this study is to provide a framework for the optimisation of a piezoelectric actuator's design based on an accurate model of the actuator, allowing to make better use of these self-sensing methodologies. An actuator morphology and its corresponding design space are first presented, a set of objective functions to optimise is then presented, after which a first pass of analytical optimisation is performed. It is found that, in order to proceed with the study, a numerical approach is required. An algorithm providing a set of optimal solutions is proposed, and is then shown to provide points from the Pareto frontier of the optimisation problem. A single optimal solution is chosen based on a tradeoff on the Pareto frontier. The resulting actuator has a sensitivity of │ß│= 3.54 nC mN−1, presenting an approximate×20 increase in sensitivity over the CMBP03 bender designed by Noliac as well as a decrease in estimation error by a factor of 50.