Short-term Power Price Forecast Based Immune Evolution Algorithm And Wavelet Neural Network

This paper studies short-term power price forecast based immune evolution algorithm and wavelet neural network. First Chinese power market and power price reform are introduced, as well as the important significance of short-term power price forecast to participants of power market. Furthermore, some methods of power price forecast are introduced as well as the characters and shortages of main methods of short-term power price forecast being analyzed. Meanwhile, this article points out that wavelet neural network shows its obvious advantages on short-term power price forecast. To solve such problems as subjectivity and dimension disaster in wavelet neural network, the paper proposes the parameters of wavelet neural network should be trained by immune evolution algorithm. By optimizing the parameters of wavelet neural network, the forecast accuracy of wavelet neural network could be improved obviously. By simulation calculation, it is found that the calculation accuracy is improved obviously as well as its adaptability enhanced by optimizing its parameters by immune evolution algorithm.