论文部分内容阅读
基于实际中常用的ＣＧＳ（ＣｌａｓｓｉｃａｌＧｒａｍ－Ｓｃｈｍｉｄｔ）、ＭＧＳ（ＭｏｄｉｆｉｅｄＧｒａｍ－Ｓｃｈｍｉｄｔ）、ＨＴ（ＨｏｕｓｅｈｏｌｄｅｒＴｒａｎｓｆｏｒｍａｔｉｏｎ）及Ｇｉｖｅｎｓ算法，给出了１类改进的直交化最小二乘新算法，分别称之为改进的ＣＧＳ、ＭＧＳ、ＭＨＴ及ＭＧＶ算法，改善了原算法的数值稳定性．将改进算法用于非线性ＮＡＲＭＡＸ模型辨识，构造出了１种新的模型结构与参数辨识的一体化算法．新算法基于逐步回归进行模型选项并消去模型中的冗余项，保证了最终模型的结构优化，并可给出比Ｂｉｌｉｎｇｓ等算法精度更高的参数估计．仿真结果证明了文章中算法的优越性
Based on the classical CGG (ClassicalGram-Schmidt), MGS (ModifiedGram-Schmidt), HT (HouseholderTransformation) and Givens algorithm, an improved class of improved least square least squares algorithm is proposed, which are called improved CGS, MGS, MHT and MGV algorithm to improve the numerical stability of the original algorithm. The improved algorithm is applied to nonlinear NARMAX model identification, and a new integrated algorithm of model structure and parameter identification is constructed. The new algorithm based on stepwise regression model option and eliminate the redundant items in the model, to ensure the structural optimization of the final model, and can be more accurate than Bilings algorithm parameters estimation. The simulation results prove the superiority of the algorithm in the article