A Multi-factor Classified Runoff Forecast Model Based on Rough Fuzzy Inference Method

Rough set theory and the fuzzy inference technique are integrated into the multi-factor medium and long-term hydrological classified forecast to solve the difficult problem of factors´ choice. The attribute reduction algorithm is used to choose the essential forecast factor sets whose classification capabilities are the best. Other forecast factor sets meeting the forecast requirement are looked after with the interactive iterative method and the concept of relative classification accuracy. The minimal decision solution is regarded as the inference rules to forecast runoff. With the case of Dahuofang reservoir in China, the establishment of the rough fuzzy forecast model is introduced and the determination process of factor sets is illuminated in detail. The results indicate that the model can effectively select the forecast factors and the forecast precision is improved.