COUPLED ALGORITHMS FOR PIEZOELECTRIC ACTUATOR DESIGN OPTIMIZATION FOR SHAPE CONTROL OF SMART STRUCTURES

Abstract

This paper presents two coupled algorithms for piezoelectric actuator design optimization for shape control of structures with both applied voltages and the shapes of actuators being treated as design variables. The optimum values for the applied voltages to actuators can be determined using the Linear Least Square (LLS) method, whereas the shapes of actuators can be optimized using the Genetic Algorithm (GA). These algorithms are combined together to develop a GA+LLS coupled algorithm, for design optimization of both actuators' shapes and voltages in either an alternating or a concurrent manner. In the alternating approach, LLS is utilized to determine the optimum voltages with given actuator geometry, and then GA is used to determine the optimal actuator shapes with given voltages; the alternating calculations continue until the selected convergence condition is met. In the concurrent approach, the LLS is embedded in GA to determine optimum voltages and then to modify the associated strings for each individual population. Numerical results are presented to validate the proposed algorithms. It is found that the concurrent GALLS algorithm appears to be most efficient and effective.