Composite system reliability analysis using particle swarm optimization

This paper presents a new approach to composite system reliability analysis. This approach is based on multi-objective particle swarm optimization (PSO), which is used as an intelligent method for scanning the state space. This paper adapts traditional binary PSO to bi-objective binary PSO using load curtailment and state probability as the two objectives to better control the particle dynamics. Another novel feature of this method is that while it utilizes information from failure states to estimate loss of load indices, it also uses information from acceptable states that are encountered to accelerate the convergence of the estimate. The method is demonstrated on the modified IEEE Reliability Test System. It is also shown to compare very favorably with Monte Carlo Simulation.