An examination of artificial immune system optimization in intelligent state space pruning for LOLP estimation

The probabilistic reliability evaluation of composite power systems is a complicated and computation intensive task. Monte Carlo Simulation (MCS) is often used as the method of choice for tackling this difficult problem, though MCS may also suffer from issues regarding high dimensionality leading to an increased need for computational resources. In order to address this issue an algorithmic method known as state space pruning has been developed in two flavors: Analytical and Metaheuristic based. The state space pruning methodology reduces the size of a given state space by removing states where there is no loss-of-load. This allows the MCS algorithm to sample a state space with a higher density of failure states which, in turn, leads to faster convergence. This study applies the CLONALG algorithm to the metaheuristic based version of state space pruning, compares and contrasts the results with genetic algorithm (GA) and particle swarm optimization (PSO) implementations, and discusses its strengths and weaknesses as applied to test systems both with and without the consideration of transmission line outages. Simulations are completed using the IEEE reliability test system (RTS) and the modified RTS (MRTS).