Abstract: Forecasted runoff is usually used to formulate a plan for short-term power generating operation of a hydropower station. However, errors in the forecasted inflow process is usually produced due to the error in runoff forecasting, and this will bring about a problem of unnecessary water abandoning or power output shortage in the case of a hydropower station operated at reservoir stage close to the limit. This paper describes a fuzzy membership function method for minimizing the effect of runoff forecasting errors based on an optimal closeness criterion to quantitatively analyze the fuzzy characteristics of the error and its influence on short-term operation of a hydropower station. And applying the credibility theory, we develop a fuzzy risk analysis model for the short-term operation, especially at reservoir stages around one of the two typical limits ? normal stage or dead stage, achieving effective coupling and transformation between the fuzziness of forecasting errors and the risk in hydropower station operation. In a case study of the Jinxi hydropower station, this model reveals the high risk operation areas, and we suggest reliable areas for its operation close to the stage limits, thus providing useful information for short-term operation of hydropower stations.

Key words: hydropower station, power generation operation, forecasting error, membership function, closeness, fuzzy risk, credibility theory