Evaluation Of Enhanced Particle Swarm Optimization Techniques For Design Of Rc Structural Elements

In this paper, the use of extended versions of basic particle swarm optimization (PSO) namely constriction factor PSO, democratic PSO and probabilistic PSO have been presented for optimal design of reinforced concrete (RC) structural elements. The design and optimization procedure follows specifications of Indian codes. Driving idea for carrying out this work was to explore extended versions of PSO for their capabilities to maximize ‘exploration area’ and minimize ‘exploration time’. These algorithms are thereby employed to study their effect on minimizing the cost of RC structural members. Optimal cross-sectional size and reinforcement for the members have been obtained by the use of computer aided environment, whereby whole process of design and optimization has been coded in C++. The design variables have been considered as continuous functions and rounded off appropriately to imbibe practical relevance to the present study. The effectiveness of these algorithms was also tested using certain benchmark functions. Various test cases of beams and columns were considered to confirm the results, and they all indicated good capabilities of these extended algorithms in terms of exploration, convergence behaviour and time. The results were also compared amongst themselves to understand adaptability of an algorithm under specific conditions.