Abstract: Reasonable operation methods help improve the power generation profits of hydropower stations. Generally, numerical simulation methods can obtain feasible solutions easily, but often have difficulty in finding out the optimal solutions. By contrast, an optimization method could obtain optimal solutions, but its given search space is often too large, thereby making the solutions unsatisfactory. Considering the characteristics of the methods of these two types, this study develops a new optimal operation method of hydropower stations based on search space reduction strategy using the equal flow regulation method. In this new method, equal flow regulation is adopted to numerically simulate the operation process, and the search space is reduced according to the process simulated. Then, in this reduced space, the operation model of hydropower stations is solved using an optimization method. A case study shows that this search space reduction method is suitable for applying the swarm intelligence algorithm to solve the operation model of hydropower stations. Under the same swarm intelligence algorithm and the same operation model of a hydropower station, it can increase the annual average power yields and firm power outputs significantly and reduce the annual average volume of waste water significantly. And in the reduced search space, the swarm intelligence algorithm converges much faster, and the standard deviations are much smaller. Thus, our new method proves effective in improving the optimization degree, convergence speed, and solution stability of swarm intelligence algorithms, very useful for the optimal operations of hydropower stations.

Key words: operation of hydropower stations, search space, equal flow regulation method, genetic algorithm, cuckoo search algorithm