Researchon Power Prediction Of Offshore Wind Farm

With the rapid development of wind energy technology, the increasing of the capacity of the individual wind turbine and the scale of offshore wind farm, the proportion is also growing in power grid.Meanwhile,wind power impact on the power grid is becoming more and more obvious. In order to satisfied the requirements of power supply, ensure the reliability and stable operation of power grid and power supply, the need for effective planning and scheduling of power system. However, offshore wind power intermittent itself unique and uncertainty, increased the power grid planning and scheduling difficulties. In order to solve the problem, the spinning reserve capacity of the power system need to be increased.In the result, the overall cost of wind power generation operation indirectly increase. Therefore, the power predict of large offshore wind farm is necessary. The accurate prediction of the short-term and medium-term of offshore wind generating capacity, can greatly reduce the power spinning reserve capacity, so as to effectively reduce the cost of wind power generation system, and provides the reliable basis for the operation of power grid dispatching.In this paper, a detailed analysis of the chaotic characteristics of time series of offshore wind power generation and unit power has accomplished.A statistical relationship between the power and single power was built. Prediction of large offshore wind farm power output parameters of each wind turbine with large offshore wind farm, greatly simplifies the complexity of mathematical model, improve the accuracy of the input data. To establish the forecast model for wind power neural network based on chaos theory, chaotic phenomena exist between the time series and the power time series of large offshore wind farm wind machine all the output prediction for offshore wind power generation, greatly improves the prediction accuracy. Medium term for the power output of the wind farm offshore prediction, proposes the prediction method of model reference adaptive capability of the neural network model of chaotic time series analysis and mesoscale weather forecast model based on the combination of the prediction precision, ensure the long time scales.