Optimal Self-identification of Adaptive Structures with Variable Geometric Parameters Using Particle Swarm Optimization Method

Abstract

The particle swarm optimization algorithm is applied to the non-linear optimization of the variable geometric parameters of the adaptive truss structure. The design variables to be optimized are the length of the truss member of the variable geometry truss, and the purpose of the optimization is to enhance the identification accuracy of the structural stiffness parameters under the concept of self-identification. The objective function in the optimization consists of multiple cost functions; a function of the condition number of coefficient matrix of linear equation for the identification and the criterion for evaluating the linear independence of the mode shapes. When the control resolution of the actuator is considered in the optimization, the discrete value of the geometric parameters is optimized by the particle swarm optimization algorithm with a penalty function. The global optimum for the continuous geometric parameters is also searched to compare the optimization results with the discrete ones. Numerical experiments on the optimization of the two-dimensional variable geometry truss are presented to show the effect of the penalty function on the value of the objective function for each mode. The deviation of the optimized value of objective function is also discussed.