Optimal Short-term Scheduling of Multiple Small-scale Hydropower Generators based on Water Inflow Prediction

The prediction of water inflow plays an vital role in the optimal operation of hydropower stations, and its accuracy is of great significance for the effectiveness of subsequent output distribution schemes. The short-term optimal scheduling is to arrange the day ahead generation plan of the power station and units under the premise of ensuring the safe and stable operation of the power grid, so as to improve the overall operation efficiency and comprehensive benefits of the hydropower stations. Based on the mined meteorological forecast data and forecast solar term data, this paper uses different neural networks to effectively predict the daily inflow of multiple units of XieCunYuan (XCY) cascade hydropower station system which is located in Songyang County, Lishui City, Zhejiang Province in the next 7 days. At the same time, the output distribution is carried out in combination with the turbine efficiency model constructed by PSO-BP and the short-term optimal scheduling model of "output determines water". Based on the operation data of XCY hydropower station in recent years, the proposed solution is evaluated experimentally. The experimental results show that Long Short-term Memory (LSTM) neural network has high accuracy in the prediction of water inflow, and the test set error is the lowest. The short-term optimal scheduling first optimizes the water inflow of the power station, and then optimizes the allocation of the water inflow of the units, which realizes the improvement of the total daily power generation and daily power generation revenue of the hydropower station.