论文部分内容阅读
本文首先介绍最小绝对误差和回归方法中的BR算法,然后根据线性规划中的灵敏度分析,对最小绝对误差和回归提出两种分别按样本点和阶进行递推的快速辨识算法,为最小绝对误差和回归的在线辨识和实时控制奠定基础。
In this paper, the least absolute error and the BR algorithm in the regression method are introduced firstly. Then, based on the sensitivity analysis in linear programming, two fast identification algorithms are proposed for the minimum absolute error and regression, respectively, according to the sample points and the order recursion. The minimum absolute error And regression online identification and real-time control.