中文摘要：

风速与风电功率预测效果一般会随着预测时间的延长逐渐变差，并使误差概率分布发生变化。比较准确地掌握风速与风电功率预测误差概率分布随时间变化的规律，对合理评估电网运行态势和优化调控电网运行策略具有重要价值。以超短期在线调度至短期预调度时窗为例，以高阶Volterra级数时间序列预测方法为工具，结合实际风电场数据，对风速与风电功率进行预测，并统计分析误差概率分布及其随预测时间变化的规律。统计发现，对特定的误差要求，当预测时长不大于60min时，风速和风电功率预测误差概率随预测时长具有基本呈线性或弱二次函数衰减的特性。

英文摘要：

In general, the forecast error of wind speed and wind power will gradually increase with the extension of forecast duration, and it makes the probabilistic distribution of forecast error varied. It is significant to master the law more accurately, by which the probabilistic distribution of forecast error of wind speed and wind power varies with time, for the reasonable assessment on power grid operation conditions and optimal adjustment and control of the operating strategy of power grid. Taking the time window covering ultra-short term online dispatch and short-term pre-dispatch as example and utilizing the forecast method based on time series founded on high-order Volterra series as the tool, based on the data of actual wind farm the wind speed and wind power are forecasted, and according to the statistics the time-varying law of probabilistic distribution of forecast error is analyzed. Statistics show that as for specific requirement to the error when the forecasted duration is ＇not longer than 60min, the probability of forecast error of wind speed and wind power appears linear characteristics or the characteristics that attenuates according to the law of weak quadratic function.