It has very important significance in the process of cascade hydropower stations optimal scheduling that give consideration to the generating capacity benefits and guarantee output power. However, considering the two indicators at same time makes the optimal operation problem of cascade hydropower stations as a multi-objective, strong coupling, complex constraints, so greatly increased the difficulty of solving. This paper put forward a kind of master-slave model self-adaptive niche genetic algorithm(MSM-SANGA), this algorithm is made by one master algorithm and several sub-algorithm, and the individual Euclidean distance criterion can self-adaptive change with the incremental evolution； finally, the MSM-SANGA is applied to solving the cascade hydropower stations multi-objective optimal scheduling problem that considers two main momentum indicators such as the maximum generating power capacity and the minimum guarantee output power etc., the calculation results show, this algorithm can be effectively applied to solving multi-objective optimization scheduling problem of cascade hydropower stations, with fast speed and high precision.