改进的零度矩阵分类算法及其在数据校正中的应用

摘要

Among the existing algorithms of data classification in material balance systems,the zero-matrix method has many advantages,such as simplicity in construction and high efficiency.However,it can′t classify data thoroughly.It will affect the performance and reliability of material balance systems.To overcome these problems,the paper proposes an improved zero-matrix algorithm which introduces the notion of absolute maximum linear independent column in the preprocessing stage of data classification.Through re-partitioning the matrix,the new algorithm can classify the data thoroughly.The numerical experiment shows that the algorithm improves the accuracy and performance of the material balance system.%在当前的物料平衡系统所用到的数据分类方法中,零度矩阵法具有构造方法简洁、运算方便、计算效率高等优点。但零度矩阵法具有对数据分类不够彻底的缺点,进而影响下一步数据校正工作的精度,最终对整个物料平衡系统的可靠性造成影响。针对上述问题,提出了一种改进的零度矩阵分类算法,在数据预处理阶段引入极大绝对线性无关列的概念,通过对矩阵的重新分割,实现了对平衡组中数据的彻底分类。仿真实验结果证明,该方法实现了对数据的彻底分类,提高了物料平衡系统的精度与可靠性。