State space pruning for power system reliability evaluation using genetic algorithms

Methods have previously been developed that improve the computational efficiency and convergence of Monte Carlo simulation (MCS) when computing the reliability indices of power systems. One of these techniques works by pruning the state space in such a manner that the MCS samples a state space that has a higher density of failure states than the original state space. This paper presents a new approach to limiting the state space sampled when calculating reliability indices by pruning the state space through the use of a genetic algorithm. This paper concludes that this technique is promising to improve the computational efficiency when calculating the loss of load probability (LOLP). This is tested using two power systems: the IEEE Reliability Test System (RTS79) and the Modified Reliability Test System (MRTS).