Investigation of the Possibility of Applying a Genetic Algorithm for Electroacoustics Problems

This paper considers the adaptation and application of a genetic algorithm to find parameters electrodynamic transducer model. The advantages disadvantages this method in comparison with classical identification using added mass are considered. derivation suitability function for estimating identified is presented, which can also be used identify other types electroacoustic transducers. theory underlying algorithms has been examined shown how work by assembling best solutions from small structural elements excellent qualities. Next, differences between traditional were analyzed, including population support use representation solutions. After that, strengths described, possibility global optimization applicability problems complex mathematical or without at all, noise resistance. Disadvantages highlighted: need special definitions settings hyperparameters, danger premature convergence. In conclusion, situations when listed not tied specific engineering scientific field, makes it universal, equally genetics computer science. determined compared more acoustics. comparative table illustrates high accuracy mass. During on practical part, improve behavior model frequencies higher than resonant, was decided complicate electrical subsystem tranducer introduce additional parameters: parallel resistance inductance. As result, complicated began correspond better measured values entire frequency domain, therefore accurate. an example convenience transition one another. results prove that appropriate solving because its allows quickly experiment models applied. Also, future, time example, nonlinear transducers state space, subject future research.This parameter selection BL fitness assessing quality directly ​​for voltage terminals converter, current through coil converter displacement moving part converter. undoubted advantage methods versatility ability adapt configure research experimentation different article describes models.