“…With the considerable progress in scientific computing power, a huge amount of industrial process data can be fully utilized, and multivariable statistical process monitoring methods have achieved remarkable results (Ge, 2017; Li et al, 2020a; Marcos et al, 2019). Among these methods, the most studied are principal component analysis (Bounoua et al, 2020), Fisher discriminant analysis (Wu et al, 2020), independent component analysis (Wu et al, 2018), and their improved versions (Guo and Zhang, 2020; Yang et al, 2020).…”