Short-term Forecasting And Optimal Operation In Tidal Power Station

Tidal energy which is a kind of clean and renewable energy has gradually received more attention as a solution for the energy shortage and environment pollution. Countries which own coastline have actively established tidal power stations want to solve energy problem. With the development of high technology, the utilization of the tidal energy has become more effective. The theory provides the basis for the large-scale development and utilization in tidal energy and the economic benefits of stations. So, the research on short-term prediction and optimal operation of the tidal power station is necessary.The tidal forecasting plays an important role in the daily operation and optimal scheduling. To tackle the non-stationary characteristics of tides caused by the non-cyclical factors, the paper proposed an improved genetic neural network prediction algorithm. First,the abnormal value in tidal data is detected and the mean substation method is adopted to overcome data error generated by the observation records. Then, the tidal forecasting model is built after optimizing node selection and resetting weights and thresholds. The application of the algorithm to actual port tide forecasting demonstrated in effectiveness.The tidal power station is similar to the common hydro-power station. The paper has researched on the optimal operation on tidal power station and established two-level economy operation model, with taking the predicted tide level and unit characteristics into account. The shortest path method is used in the first level while the PSO algorithm is used in the second.At last, this paper proceeds the calculation of tidal power station. The result shows that the optimal regulation can improve the electricity generation effectively.