An improved transfer learning approach based on geodesic flow kernel for multiphase batch process soft sensor modeling

Abstract

For multiphase batch process, the characteristics of process data under various batches differ. Consequently, the soft sensor model built for a particular working condition is inapplicable to other working conditions. Besides, each batch can be divided into several phases whose characteristics are probably different. To address the above problems, a soft sensor model based on phase division and transfer learning strategy is proposed. First, transfer learning strategy is adopted to construct a soft sensor model applicable to various working conditions. Specifically, geodesic flow kernel based on linear local tangent space alignment (LLTSA-GFK) algorithm is designed. By projecting process data to the common manifold subspace, the distribution difference of data between various batches is reduced and the performance of the soft sensor model is enhanced. In addition, sequence-based fuzzy clustering and just-in-time learning (JITL) are adopted to solve the multistage characteristic for batch process. The root-mean-square error ( RMSE), coefficient of determination [Formula: see text], and mean absolute error ( MAE) are adopted to compare the conventional soft sensing approach (i.e., partial least-square regression based on JITL, support vector regression, and back propagation neural network) with the proposed approach. The superiority of the proposed model is verified by a fed-batch penicillin fermentation process.

Modeling and Optimization of an Enhanced Soft Sensor for the Fermentation Process of Pichia pastoris

“…However, as industrial demands evolve, the working conditions for the fermentation process of Pichia pastoris are subject to change, leading to a noticeable increase in the predictive error of conventional models. Consequently, researchers have proposed soft sensor modeling methods based on transfer learning, which have proven effective in mitigating model errors amidst changing conditions [9][10][11]28,29]. Further, this paper hypothesizes that with limited labeled samples available from a single-source working condition, knowledge transfer from just one source is insufficient to effectively enhance predictive performance under target conditions.…”