The core value of grain-oriented silicon steel as a functional material in the field of power transmission and distribution and its important role in energy conservation and emission reduction were expounded. In addition, the carbon emission situation of grain-oriented silicon steel and its transformer in the whole life cycle were quantitatively analyzed through data study. Finally, according to the quantitative analysis results, the development direction of Chinese grain-oriented silicon steel industry was inferred. 

Grain-oriented silicon steel is a kind of iron-silicon binary alloy widely used in electrical equipment manufacturing industry. It needs special properties such as low iron loss, high magnetic induction intensity, low magnetic aging, good punching performance and high surface quality. Its production processes and equipments are complex, and its chemical composition is strictly controlled. According to the production processes and performances, it is mainly divided into ordinary grain-oriented silicon steel (CGO) and high magnetic induction grain-oriented silicon steel (HiB). Starting from the product requirements of grain-oriented silicon steel, the latest production progresses of domestic and foreign enterprises such as Japan's Kawasaki Steel and Nippon Steel, China's Baowu Steel and Shougang Corporation were introduced, and the technological processes and quality control points of this type of steel  were discussed, and the key production technology of grain-oriented silicon steel were  emphatically introduced. Finally, six new processes for producing grain-oriented silicon steel, including thin slab continuous casting and rolling, asynchronous rolling, magnetic domain refinement, production without traditional inhibitor, double grain-oriented silicon steel and magnetic field annealing, were discussed. The new technologies of silicon steel looked forward to the development trend of grain-oriented silicon steel.

The effects of normalizing treatment on microstructure, texture evolution and magnetic properties of  3.0 % Si non-oriented silicon steel after hot rolling were investigated in laboratory. The results showed that the microstructure of the hot rolled plate was completely recrystallized after normalizing at  900 ℃ , and the average grain size increased with the normalizing temperature rising to  1 000 ℃ . The average grain size was  222 μm , and the microstructure uniformity was obviously improved. When the normalizing temperature increased from  900 ℃  to  1 000 ℃ , the magnetic induction  B50 increased from 1.661 T to 1.674 T. Low frequency iron loss  P1.5/50 decreased from 2.323 W/kg to 2.282 W/kg, and high frequency iron loss P1.0/400 decreased from 18.37 W/kg to 17.45 W/kg.

Based on the desulfurization theory and actual production data analysis, the pretreatment desulfurization of silicon steel in a factory was studied. The results showed that the sulfur content of the secondary desulfurization slag was proportional to the sulfur content in the front slag,inversely proportional to the front slag residue, proportional to the primary desulfurization slag residue, and inversely proportional to the amount of desulfurizer. When the total amount of desulfurizer was constant, the sulfur content in the secondary desulfurization slag decreased first and then increased with the increase of the proportion of primary desulfurizer, and the best proportion of primary desulfurizer was  40 % . With the increase of temperature before pretreatment,the sulfur content of molten iron decreased first and then increased, and the optimum temperature before pretreatment was 1 370-1 400 ℃. The sulfur content of hot metal was reduced to 11.7×10^-6 by three times of slag removal and two times of desulfurization, which met the production demand of silicon steel.

The PLS (partial least squares) and BP network were applied to establish the online iron loss prediction mathematical model of silicon steel. The results showed that PLS and BP network have higher fitting and prediction accuracy, and BP network has higher accuracy. The model can be applied to on-line prediction of iron loss of silicon steel and development of new grade silicon steel products. 

A texture data analysis method based on Matlab was introduced, through the secondary programming of MTEX, an open source project of Matlab, realizing the data analysis function of silicon steel texture. It can break away from the limitations of equipment, software and space, and conveniently deal with EBSD and XRD texture data, including data noise reduction, drawing (BC, IPF, PF, ODF, KAM), calculation of texture proportion, average grain size and grain size distribution. 

The causes of the interlaminar misalignment of nonoriented silicon steel finished coils of the finishing unit were analyzed, focusing on the causes of the lateral deflection of the steel strip in the uncoiling area. The interlaminar misalignment has reduced effectively by improving the equipment accuracy in the uncoiling area, modifying the key equipment, and optimizing the production operation mode, which then improved the coiling quality. 

A method for improving the structure of brushes in cleaning section of silicon steel annealing line was revealed in this paper. To configure the length of brush roller bristle area by gradient, and single-point control brush roller ascending and descending movement, the cleaning problem of steel strip rolling oil was effectively solved, which significantly improved the cleaning quality of sheet surface, reduced the defects of the final products, and meanwhile reduced the cost of production.

Mibach laser welding machine is one of the important equipment of the normalization unit. The welding quality determined by the shearing effect is very important to the operation of the whole unit. By optimizing the adjustment method of the cutting box device of the welding machine, the cutting effect was greatly improved, which guaranteed the weld quality and improved the overall operation rate of the unit.