基于决策树的风电机组偏航启停优化研究

Research on optimization of yaw start and stop for wind turbines based on decision tree

风速和风向的频繁变化会导致偏航系统频繁启停,增加场用电。选取风速较低时,风电机组偏航系统的运行状态进行研究,通过分析风速、风向、有功功率、机舱位置和偏航误差,进一步得到每个监测点的变化。由于风速和机舱位置是影响偏航系统启停的重要参数,因此,需要考虑上一个监测点的风速变化和机舱位置的变化。建立基于CHAID的决策树模型,通过决策树分类偏航相关的参数,然后,运用这些参数分析低风速下机组的偏航启停的历史数据,依据决策树的分类规则在线优化机组低风速状态下的偏航系统的启停。通过实例验证了优化后机组的偏航次数和偏航时间减少了。该研究为偏航系统的运行优化和提高偏航系统的可靠性奠定了基础。

Due to frequent changes in wind speed and wind direction, the yaw system frequently starts and stops, increasing power consumption. The running state of yaw system of wind turbine is studied when the wind speed is low. By analyzing wind speed, wind direction,active power,nacelle position and yaw error, the changes of each monitoring point are obtained. Since wind speed and nacelle position are important parameters that affect the start and stop of the yaw system, it is necessary to consider changes in wind speed and nacelle position at each monitoring point. The yaw related parameters are classified according to the decision tree. These parameters are used to classify historical data on the operation and shutdown of the yaw system of the unit at low wind speeds. According to the classification rules of the decision tree, the yaw system start and stop under low wind speed conditions are optimized online. After optimization, the yaw frequency and yaw time of the unit are reduced. This study laid the foundation for optimizing the operation of the yaw system and improving the reliability of the yaw system.