论文部分内容阅读
对滑坡预测预报的非线性模型,在估计其参数时,传统的方法是将非线性模型在参数的近似值处展开成泰勒级数,并仅取至一次项,然后再应用线性模型参数估计理论进行参数估计。因线性化时略去了二阶及二阶以上的各高次项,所以必然会产生模型误差。介绍了高斯—牛顿法的基本原理,并以洒勒山新滑坡为例,在建立该滑坡灰色GM(1,1)模型和Verhulst模型的基础上,运用高斯—牛顿法对两个非线性模型的参数进行优化。计算结果表明,参数优化后各模型的预测精度比优化前各模型的精度有显著提高。说明采用高斯—牛顿法优化非线性模型参数是提高滑坡预测预报精度的一种有效且切实可行的方法。
In the nonlinear model predicting landslide, when estimating its parameters, the traditional method is to develop the nonlinear model into Taylor series at the approximation of the parameters and take it only once, and then apply the linear model parameter estimation theory Parameter Estimation. Due to the linearization of the second and second order omitted more than the high-order items, it will inevitably produce model errors. The basic principle of Gauss-Newton method is introduced. Taking the new Spalingshan landslide as an example, on the basis of establishing gray GM (1,1) model and Verhulst model of landslide, Gaussian-Newton method is applied to two nonlinear models Optimization of the parameters. The calculation results show that the prediction accuracy of each model after parameter optimization is significantly higher than the accuracy of each model before optimization. It shows that using Gauss-Newton method to optimize nonlinear model parameters is an effective and practical method to improve the accuracy of landslide prediction.