An Efficient Optimization Design for 1 MHz Ultrasonic Transmitting Transducer

Abstract

Due to the low efficiency of traditional design method, an efficient optimization design method is developed to design and fabricate high-performance ultrasonic transmitting transducer (UTT) based on PiezoCAD software and particle swarm optimization (PSO) algorithm. Based on the data of design and performance parameters of UTT obtained by PiezoCAD software, the genetic algorithm-based back-propagation neural networks (GA-BPNNs) are established to describe their mapping relationship. The optimality criterions are established based on the performance parameters including center frequency (CF) and maximum echo amplitude (MEA). Then, based on the established GA-BPNNs and optimality criterions, the optimization design method of UTT is developed under the framework of PSO algorithm. According to the desired performance, the design parameters of UTT, including the thickness and diameter of PZT-4 (Pb(Zr x Ti1-x)O3), the acoustic impedance of backing, are optimized by the developed method, and they are 2 mm, 30 mm and 3 MRayl, respectively. Based on the optimized design parameters, the CF and MEA simulated by PiezoCAD are 1.05 MHz and -45.1 dB, which are in good agreement with the designed ones (1 MHz and -45 dB). In addition, the CF of the fabricated UTT is 1.07 MHz, and its insertion loss is -7.2 dB at 1.075MHz, which implies that the fabricated UTT has excellent performance. Therefore, the developed optimization design method is effective to fabricate high-performance UTT.