分段最小二乘（Segmented Least Squares） 否则，i 与之后一直到 j 所有的点拟合一条新的线的cost + 新增加了一条线就会多出的代价C + i 点之前的最小平方和OPT(i-1) .
基于2个网页-相关网页
该算法是在最小二乘意义上的多项式曲线拟合，实验结果表明，其校正效果明显优于分段线性校正算法。
The algorithm is polynomial curve fitting based on least square approximation. The experimental result shows its performance is more effect than that of the piecewise linear algorithm.
采用基于最小二乘的分段多项式，建立了不同温度下EGO传感器的静态模型。
The static models of EGO sensors for different temperature were built up using sectioned polynomial based on the least square.
基于区间分段思想，将极化曲线非线性参数辨识转化为两个线性最小二乘辨识子问题。
The nonlinear parameter identification of fuel cell polarization curve was converted into two sub-problems of linear least square identification based on interval segmentation technique.
应用推荐