Unit Commitment Involving Uncertainty in Electric Power Systems Using Orthogonal Experiments

Unit Commitment (UC) in power systems involves determining the optimal operation schedule of power-generating units to meet forecasted electricity demand while adhering to various operational constraints. The inherent variability in demand forecasts and the intermittent nature of renewable energy sources (such as wind and solar) introduce significant uncertainty into the UC problem, necessitating advanced algorithms capable of handling these uncertainties. This paper addresses UC studies involving uncertainties due to renewable power generation and load fluctuations using orthogonal experiments. Statistical percentiles within cumulative distribution functions and perturbations in renewable power generation and loads are utilized to quantify two levels of uncertainty. Various orthogonal scenarios, based on the Taguchi method, are analyzed to identify individual UC solutions, taking into account these two levels of uncertainty, obtained by the point-estimation method or percentiles. The proposed method’s applicability is demonstrated using both a 10-generator system and a realistic 52-generator system with a peak load of 39.718 GW, incorporating renewable power generation. The results indicate that the orthogonal scenarios combined with the point-estimation method yield an operational cost closest to the benchmark cost obtained from the Monte Carlo simulation. Additionally, this approach exhibits the smallest standard deviation in costs, closely matching that of the Monte Carlo simulation. Furthermore, the proposed method requires significantly less CPU time compared to both the traditional point-estimation method and the Monte Carlo simulation.