Electricity power is a basic industrial component which plays an important role in the economy of a nation. In this paper, the correlations evolution of electricity prices among 50 states and the District of Columbia are studied based on random matrix theory (RMT) Four regime shifts are identified from January 1990 to August 2014 in the U.S. residential, commercial and industrial electricity markets. Then, the genetic algorithm (GA) is applied to analyze the clusters of evolution. The results show that, the correlations of electricity prices increased continually in the three departments. However, it decreased in 2012 which further confirms its sensitivity to fuel market. Besides, four regime shifts exist in the three departments though the different times of occurrence caused by price level. And, the fluctuation of community evolution is consistent with four regime shifts. The final part is a summary of the research analyzed and results. (C) 2016 Elsevier Ltd. All rights reserved.