带权优化方法在曲线预测模型中的应用研究

Application of the Weight-carrying Optimistic Method on the Curve Predictive Model

为了能够根据监测数据推测被测量变化规律及准确预测未来变化,针对常规最小二乘法拟合精度较差的情况,提出利用带权优化方法在常规非线性拟合基础上进行优化处理。通过对工程实际监测数据的计算验证表明,带权优化得到的拟合曲线,拟合精度与预测效果均有明显提高。

Because of the lower accuracy in the general least squares fitting, the weight-carrying optimistic method is used on the optimistic processing of data on the basis of the common nonlinear matching so as to estimate the measurand variance and make a perfect forecast of future changes from the monitoring data. The calculation of the actual monitoring data obtained from an engineering shows that it can greatly improve the fitting accuracy, predictive validity and matched curves gotten from the weight-carrying optimization.