Abstract:

Aim ing at the curse of dimensionality of power generation dispatching in cascade reservoirs with hybrid pumped storage power stations,based on the randomness and nonlinear characteristics of historical runoff series,a longterm optimal dispatching model of cascade reservoirs with hybrid pumped storage power stations was established.In the scheduling period,the maximum generating capacity of the cascade pumped storage power station was selected as the objective function,and the improved social emotional optimization algorithm was introduced to solve the problem.Taking a mixed flow cascade power station as an example,the simulation results showed that the new method could reduce the redundant calculation of traditional algorithms,and effectively combined the decision-maker’s subjective preference as well as interest propensity in the optimal operation of cascade type reservoirs in hybrid pumped storage power stations. Compared with the dynamic programming algorithm,the new method used a high-quality solution with an annual power generation of 66.96×10⁸ kWh in a short period of time.

Key words: hybrid storage hydropower station, reservoir operation, social emotional optimization algorithm, dynamic programming, electricity generation, cascade reservoir