汽油机热怠速排放实验数据的符号化分析

Symbolization Analysis of Experimental Emission Data of Gasoline Engine at Hot Idle Speed

介绍了符号时间序列分析中符号的转换、直方图的绘制、符号树的结构及其信息熵;选择了符号的数目、固定字长度、划分线的位置、采样时间间隔等参数,并将符号时间序列的方法应用于汽油机热怠速排放实验数据.用MATLAB编程,得到了符号序列直方图,求出了Shannon熵、欧几里得范数等统计量,并对其结果进行了分析,肯定了统计量对确定性结构的评估作用.

This article introduces symbol transformation and how to plot symbol-sequence histograms as well as the structure of symbolic tree and Shannon entropy. Parameters such as symbol-sequence length and symbol-set size and modified Shannon entropy etc are selected by considering many factors. Experimental data emitted by gasoline engine at hot idle speed is analyzed in symbol time series method in this article. And symbol-sequence histogram are got by programmed in MATLAB. Statistics such as modified Shannon entropy and the Euclidean norm are extracted too. An the end,all results are explained. It proposes that Shannon entropy can evaluate deterministic structure in symbol-sequence histogram.