基于梯度遗传算法压电换能器稳健优化设计.pdfVIP

256 Vol. 39 No.2 2011 2 Computer Digital Engineering 156 * ( 621010) , , , , , , , , , ; ; ; N384; B472 Robust Optimization Design for Cantilever-type Piezoelectric Vibration Transducer Using Gradient-Assisted Genetic Algorit m Yu Yali Yang ao (School of Information Engineering, Southwest University of Science and echnology, M ianyang 621010) Abst ract Cantilever-type piezoelectric vibration transducers are a new class of pow er supply devices w ith a variety of uncertain fact ors, especially manufacturing errors. Based on aguchi robust design, genet ic algorithm and the theoretical model of cant ilever-type piezoelectric vibration transducers, the optimization method of robust design w ith average pow er out- put as the robust opt imization objectives is developed. he convergence speed of genetic algorithm is accelerated by the grad-i ent informat ion of object ive function. Each generation of population generated by genetic algorithm is used as a group of data in the inner tables. A method with crossing inner and outer tables is adopted to comput e the SNRs ( Signal Noise Ratios) of optimizat ion objectives. hrough analyzing the SNRs, a set of optimal design parameters are obtained. Simulation results show transducers with optimal paramet ers exhibit satisfactory charact eristics. Key Words piezoelectric vibration transducer, robust design, gradient, genetic algorithm Class Number N384; B472 1 , , , [ 1] , [ 3] ,