Taking Ankang reservoir as the research object,this paper investigated the effects of hydrological sequence variation on the results of reservoir operation. It used 3 methods such as Mann-Kendall rank test method, sequential clustering and the anomalies accumulate method to diagnosis the runoff series in Ankang reservoir from 1950 to 2014 so as to determine the aberrance point and established the model of energy maximization,used Genetic Algorithm optimization to solve the reservoir operation results of two inflow runoff series before the aberrance point (serious 1) and after aberrance point (serious 2). The results show that: the runoff in Ankang reservoir occurred variation in 1984, and then water yield of the variation sequences showed a decreasing trend; compared with the serious 1, the average electricity generation and the overall water level of Ankang reservoir has a decreasing trend in serious 2, but the water level is opposite in peach flood season. The result has great practical value for the design of hydropower station and the operation of reservoir and so on.