Efficient parallel design/placement optimization methodology of large smart structures

摘要

The tide methodology is formulated for optimal sizing of structural members and/or placement of active elements for structural weight minimization and shape control of piezoelectric-type smart structures in a parallel environment The efficient,stochastic-type Integral Global/Local Optimization algorithm (IGLOA) is first extended to form a discrete design variable (D) version, IGLOA_D, and used to develop optimization procedures to solve this class of mixed integer/discrete/ continuous designvariable optimization problems. The design variables include both structural sizing (such as cross-sectional area, etc.) and active element placement/design (such as number, location, and/or control gain) variables. Several numerical design/placementoptimization examples of 72- and 504-bar piezoelectric-type smart truss structures are presented for demonstration purposes. Excellent computational timing and optimization results have been obtained, indicating that the algorithm/methodology is capableof solving this class of computationally intense, large-size, smart structural optimization problems.