Cascade Hydropower Stations Monthly Runoff Forecast Algorithms

Economic dispatching of series hydroplant system plays an important role for optimal operation of power grid. Appropriate scheduling of the series hydroplants is vital to the security, stability and economic operation of the whole power system.The hydrologic forecasting is the basis of the economic dispatching of multi-reservoir system. Precise prediction is the key to ensure the success of optimal dispatching. This thesis, based on the practical engineering background of cascaded reservoir along Wasi river, builds a hydrologic forecasting model which can reflect the actual condition of the real operation process, with detailed analysis, generalization and idealization.After analyzing the different kinds of forecasting models, four kinds of forecasting model are proposed, including local modeling, grey forecasting model, BP neural network and BP neural network based on local modeling. The models are simulated under Matlab? environment. The simulation result shows that the first three kinds of algorithms cannot achieve accurate result, producing large error on the data that are out of the historic record. While the algorithm BP-NN based on local modeling can obtain comparatively precise results and has higher predicting capability for hydrologic forecasting. This algorithm has strong adaptability, usability and overall performance. Hence it might be a feasible model for further research on the prediction of the complicated non-linear hydraulic systems.