基于黄金分割法的离散回归方程的变点估计及其应用

Change-Point Estimation for Discrete Regressive Equation Based on the Golden Section Method and its Application

本文利用变点统计学和黄金分割法讨论有多个变点的离散回归方程的交点估计和参数估计，文中提出基于黄金分割法搜索最佳变点估计和同时得到参数估计的最小二乘算法，还讨论该算法在控制领域的应用，数值模拟结果显示本文算法能给出良好的变点及参数的估计值。

The ideas of change-point statistics and the golden section method are applied to the change-point estimation and parameter estimation problems for discrete regressive equation with several change -points in this paper. The searching method of optimal change- point estimation based on the golden section method is discussed in detail and then the least squares algorithms for solving the change- point estimation and parameter estimation are given.In theis study. foe application of foe methods to control fields and its computer numerical simuiation are discussed. and the simuiation results show us the good estimation both of change-points and parameters by the methods