夹杂物尺寸及数量对无取向硅钢磁性能影响的主成分回归分析

Principal component regression analysis on the size and quantity of inclusions affecting magnetic properties of non-oriented silicon steel

采用扫描电镜、场发射扫描电镜、能谱仪等对50SW1300冷轧无取向硅钢中的夹杂物分不同尺寸区间进行数量统计,利用主成分回归分析法,即数据的标准化处理—主成分分析—回归分析—标准化的变量还原成原始变量—确定显著影响因素,综合分析夹杂物总量及各尺寸区间的夹杂物数量对无取向硅钢磁性能的影响。结果表明:主成分回归分析能够从夹杂物尺寸区间及数量的多个影响因素中提取主要的因素,定量研究其对磁性能的影响。分析表明,显著影响无取向硅钢铁损的夹杂物为100~500nm的AlN、AlN+MnS、MnS、Al2O3、AlN+Al2O3,而劣化磁感最明显的夹杂物尺寸区间为100~200nm。

Different size intervals of inclusions in cold rolled non-oriented silicon steel 50SW1300 were counted by scanning electron microscope(SEM),field emission scanning electron microscope(FESEM)and energy disperse spectroscopy(EDS).With principal component regression method:standardization for experimental data,principal component analysis,regression analysis,transform standardized variables into original variables,determination of significant factor,effects of the total number of inclusions and the quantity of each size intervals of inclusions on magnetic properties of non-oriented silicon steel were comprehensively analyzed.The results show that the main factors can be extracted by principal component regression from multi-factors such as the total number of inclusions and the quantity of each size intervals of inclusions,which used to quantitatively research the influence on magnetic properties.It is indicated that the inclusions which significantly affect the core loss of non-oriented silicon steel are AlN,AlN+MnS,MnS,Al2O3,AlN+Al2O3and the size interval is 100to500 nm,and the inclusions in the range of 100 to 200nm reduce the magnetic induction greatly.