Improved fast short-term wind power prediction model based on superposition of predicted error

Accurate prediction of short-term wind power is an effective way to rationally adjust the scheduling strategies and to improve the operation stability and economy of microgrid with wind turbines. A new improved prediction method which does not rely on any basic prediction methods is proposed based on analysis of a traditional wind power prediction procedure and in terms of strategy how to use a basic prediction method in wind power prediction procedure. By the proposed method, an additional error prediction model is built to predict the error of the predicted results by the traditional prediction method. And the predicted error value is added back to the predicted results mentioned above to reach the final predicted results. Taking Back propagation (BP) neural network as a basic prediction method, the proposed prediction method is validated by output power prediction of a real wind farm. Support vector machine (SVM) is chosen as another basic prediction method to test the versatility of the proposed improved prediction method. The simulation results show that the proposed wind power prediction method can improve the prediction accuracy by about eight percent. The proposed method does not involve the internal characteristics of any basic prediction methods or require any auxiliary methods, which is quite different from the traditional improved methods available and thus is more universal.