提高风电功率预测精度关键技术研究

Accurate estimation and real-time analysis of highly uncertain and intermittent characteristics of wind power are the urgent problem that wind power transportation needs to solve. It is one of the important means to solve this problem that wind power forecast. For a large error of current numerical weather forecast and the low precision problem of wind power prediction with the model forecast and physical method in some wind farms, this project is to develop by modeling and experimental research. Multi-dimensional flux form fluid mechanics and thermodynamic control equations are established based on the numerical model framework of WRF medium medium-mesoscale weather forecast, with combining theories of physics, meteorology and thermodynamics. Plug-in software package of a small-medium- mesoscale (1-5km) numerical weather forecast model is constructed, with the initial field, boundary field, grid number and other calculation modes refined. The change theories of wind energy are acquired in the near-surface layer, with exploring the mechanisms of meteorological information affected in the wind field. The output revision models of the numerical weather forecast and the wind power statistics prediction models are established, with constructing wind farm parameter models that can be selected such as terrain, landform, contour, roughness, obstacles, temperature stratification and humidity stratification, and combining observation data of wind towers around and the actual output during the same period in the wind farms. After sensitivity tests of the prediction systems are conducted, wind power prediction models are quantified and determined, and wind power forecasting systems are improved in wind farms, the key technical problems of wind power prediction accuracy improved may be achieved.

风电不确定性和间歇特性的准确估计与实时分析是风电入网迫切需要解决的难题，风电功率预测是解决该问题的重要手段之一。本项目拟在前期研究的基础上，针对当前数值天气预报在某些风电场存在较大误差，模式预报结合物理方法预报风电功率精度不高的现实问题展开建模与实验研究。基于ＷＲＦ中尺度天气预报数值模式框架，结合物理学、气象学和热力学等理论，建立多维通量形式流体力学和热力学控制方程组，细化初始场、边界场和格点数等计算模式，构建中小尺度（1-5km）数值天气预报模式外挂软件包。探讨影响风场气象信息机理，获得近地层风能变化规律，构建可自选地形、地貌、等高线、粗糙度、障碍物、温度分层和湿度分层等风电场参数模型，结合风电场周围测风塔观测资料和同期的风电场实际出力，建立模式输出订正模型和风电功率多元统计预报模型。开展预报系统敏感性试验，量化风电场功率预报模型，完善风电场功率预报系统，解决提高预报精度的关键技术难题。

风电不确定性和间歇特性的准确估计与实时分析是风电入网迫切需要解决的难题，风电功率预测是解决该问题的重要手段之一。本项目针对当前数值天气预报在某些风电场存在较大误差，模式预报结合物理方法预报风电功率精度不高的现实问题展开建模与实验研究。基于ＷＲＦ中尺度天气预报数值模式框架，结合物理学、气象学和热力学等理论，建立多维通量形式流体力学和热力学控制方程组，细化初始场、边界场和格点数等计算模式，构建中小尺度（1-5km）数值天气预报模式外挂软件包。探讨影响风场气象信息机理，获得近地层风能变化规律，构建可自选地形、地貌、等高线、粗糙度、障碍物、温度分层和湿度分层等风电场参数模型，结合风电场周围测风塔观测资料和同期的风电场实际出力，建立模式输出订正模型和风电功率多元统计预报模型。开展预报系统敏感性试验，量化风电场功率预报模型，完善风电场功率预报系统，解决提高预报精度的关键技术难题。安西、酒泉风电场应用表明：基于3S技术的风力发电智能调度系统、基于WRF和3S技术的风电功率智能预测系统、基于WRF的风电场风资源智能预测系统能满足安西、酒泉风电场应用要求。短期风电功率预报误差占装机容量的百分比低于10%，0-4小时超短期风电功率预报模型的误差小于3%，中长期预测误差小于5%，有效提高风电功率预报精度，具有实际应用意义，可以推广。研制的激光风速风向仪，风速风向遥感探测器《离心分离器》和风电储能设备样机能满足工业生产需要，具有一定推广价值。