二次曲线指数平滑法应用在预测中,仅凭人工经验选取平滑系数所得的预测结果并不准确。针对风电场监测预警中的实际问题,提出采用缩小平滑系数选取区间,得出最优平滑系数的二次曲线指数平滑法,建立杆塔沉降预警模型,并通过该模型对风电机杆塔沉降度进行预测。改进平滑系数后的二次曲线指数平滑法在预测的平滑度和信度上,都有非常明显的优化效果,从而更准确地预知数据异常所构成的潜在危险。
When the quadratic curve exponential smoothing method is used in the prediction, the predicted results are not accurate if the smoothing coefficient is selected just by artificial experience. Aiming at the practical problems in wind farm monitoring and early warning, this paper proposes to narrow the smoothing coefficient selection range and get quadratic curve exponential smoothing method of the optimal smoothing coefficient, establishes the tower settlement early-warning model, and forecasts the settlement degrees of wind turbine tower through this warning model. The quadratic curve exponential smoothing method after improving smooth coefficient has obvious optimization in the prediction of smoothness and reliability, so as to more accurately predict potential dangers caused by abnormal data.