Genetic algorithm-based integrated optimization of active control systems for civil structures subjected to random seismic excitations

Abstract

An efficient genetic algorithm-based approach is developed to solve the integrated optimization problem for active control systems of civil structures under random seismic excitations. Not only the number and location of actuators but also the control algorithm can be optimized simultaneously by a novel genetic algorithm while the randomness of seismic excitations is taken into account. In view of the randomness of earthquake actions, optimal control criteria should be established in a probabilistic sense. Considering the trade-off between control effectiveness and energy consumption, two sets of probabilistic criteria for stochastic optimal control of structures are therefore developed, on the basis of peak and root-mean-square values of objective quantities. In the proposed genetic algorithm, an effective selection policy, an improved binary single-point crossover and a new single-locus or two-locus mutation are adopted. Finally, the effectiveness, efficiency and stability of the proposed optimization methodology are verified by two numerical examples.