Data-driven approach of FS-SKPLS monitoring with application to wastewater treatment process

Abstract

In this paper, a data-driven scheme of spherical kernel partial least squares based on feature subspace (FS-SKPLS) will be applied to the wastewater treatment process (WWTP). First, select appropriate data variables. Utilize the benchmark simulation model no. 1 (BSM1) to obtain large amounts of training and testing data needed in the process monitoring. Then, introduce the feather subspace method into spherical kernel partial least squares (SKPLS) to get FS-SKPLS approach. Finally, adopt FS-SKPLS to achieve the aims of the off-line modeling and the on-line process monitoring.