The Evaluation of Resonance Frequency for Piezoelectric Transducers by Machine Learning Methods

Abstract

Piezoelectric transducers are used in the application of sending and receiving sound waves. The distance at which the sound waves can be transmitted is determined by the transducer frequency. Therefore, measuring the frequency becomes an important issue. However, a large number of experiments are needed in the laboratory to simulate and measure the resonant frequency of the transducer. To simplify this problem, this study uses machine learning methods rather than laboratory experiments to estimate the frequency of transducers. The proposed method is compared with other methods, such as an artificial neural network, support vector machine, neuro-fuzzy, and mega-fuzzification. The results show that machine learning methods are efficient ways to assess the resonance frequency of a piezoelectric transducer, and mega-fuzzification method has the best accuracy among the comparative methods in this case.