In order to reduce the effect of reservoir operation on river ecological system and solve the ecological problems of Nilki reservoir, an effective ecological scheduling scheme was developed combined with the hydrological variation diagnosis system． Firstly, the hydrological time series was analyzed by hydrological diagnosis method to determine the time node of different human disturbances. Meanwhile, considering the icebound season characteristics in Northeast China, the natural ecological, poor natural and current runoff processes under the mixed time scale were calculated. Secondly, the Nilki reservoir ecological scheduling model restrained by multiple ecological runoff process was established aiming at maximizing the power generation capacity. Finally, according to the water balance equation, the initial population was generated by normalized method to restrict the variables to the same perception radius, and the scheduling process under five kinds of constraints was optimized by firefly algorithm. The results showed that the original river ecological system was not protected effectively under the existed dispatching rules, water was not reasonably used in wet year. Considering the hydrological variation of the ecological scheduling, a low ecological assurance level scheduling scheme was provided for decision-makers by reducing ecological objectives in dry year to reduce the damage to the river ecosystem. The contradiction between hydropower generation and ecological objectives was more obvious in dry season, and it was effectively alleviated in wet season. Therefore, decision-makers need to make scientific decisions in different periods.