基于混沌—LSSVM神经网络风电场风速短期预测

摘要

Wind power output is the basic data on which electric power system operation and planning are based.Accurate prediction of the wind speed is helpful to improve the economy and reliability of the power system in operation.In this paper,under the premise of the wind speed time series with chaos characteristics,combined with the chaotic time series of phase space reconstruction and with the support of vector machine (SVM) regression theory,a short-term wind speed prediction model is set up based on wind speed chaos characteristics and the most popular least squares vector machine.The hybrid algorithm is used for wind speed forecast of a wind farm in zhangjiakou.Through the simulation calculation and analysis,it has shown that the hybrid algorithm-chaos-LSSVM neural network can further increase the prediction accuracy.%风电出力是电力系统运行与规划的依据,准确的风速预测有利于提高电力系统运行的经济性和可靠性.基于风速时间序列具有混沌特性的前提下,结合混沌时间序列的相空间重构和支持向量机回归理论,建立了一种基于风速混沌特性和当前最为流行的最小二乘向量机的短期风速预测模型.用于张家口某风电场进行风速预测,通过实例仿真计算分析表明,混沌—LSSVM神经网络的混合算法可进一步提高预测精度.