Weighted differential evolution heuristics for improved multilayer piezoelectric transducer design

Abstract

Piezoelectric transducers are designed for wide applications of underwater sonar, energy harvesting, biomedical and nondestructive testing in terms of higher sensitivity and broad bandwidth. Multilayer transducers have well-known advantages of higher sensitivity and thus can be optimized by varying layer thicknesses of active material. The present study deals with mathematical optimization of multilayer transducers exploiting weighted differential evolution (WDE) based computational heuristics for optimized piezoelectric transducer design while satisfying the constraint of constant total thickness. Fitness evaluation is carried out on the basis of one dimensional model (ODM) for simple multilayer, multilayer with bondlines and transducer with bondlines and matching for piezoelectric materials including PZT5h, PMN-PT and piezocomposite with PMN-PT for different loading media. WDE exploited along with ODM for the transducer models has demonstrated the efficacy, reliability and applicability of the proposed scheme. Comparative study of the proposed WDE-ODM with its counterparts including simulated annealing, genetic algorithms, and particle swarm optimization, reveal better performance in terms of accuracy, complexity and convergence metrics for different number of layers for simple transducers. Applicability of WDE-ODM on piezoelectric structures with bondlines, and matching layers has also been accessed and validated.