Soft sensor development based on improved just‐in‐time learning and relevant vector machine for batch processes

Abstract

The online measurement of key quality variables based on soft sensors plays a critical role in ensuring the safety and stability of batch processes. Recently, the relevant vector machine (RVM) was introduced into soft sensors for batch processes. However, the RVM-based soft sensor has limitations in addressing the time-varying, high-dimensional, and dynamic data of batch processes. To address these issues, based on improved just-in-time learning and the relevant vector machine, an adaptive soft sensor, termed IJITL-RVM, is proposed in this paper. The IJITL-RVM integrates the IJITL algorithm and the RVM algorithm into a unified online modelling framework with the ability to perform adaptive updating and dynamic modelling. First, to enhance the performance of online prediction, an IJITL is designed to select modelling data based on the support vector data description (SVDD) algorithm and the kernel trick. Based on the comprehensive consideration of the strong nonlinearity and high dimensionality of process data, the IJITL can adaptively and accurately select the modelling data. Afterward, a local model is established by using the RVM for online prediction. Three applications, including a numerical simulation example, some UCI datasets, and a penicillin fermentation process, are provided to illustrate the superiority of the IJITL-RVM-based soft sensor.

A review of just‐in‐time learning‐based soft sensor in industrial process

“…Kim et al [82] developed an inferential control system that integrates a commercial model predictive control software and an LW-PLS soft sensor. GPR [110,111,112,113,114] GMR [115] JITL based neural network [117,118,119,48,120] Semi-supervised strategy [121,122,123] Dynamic modeling [124,125] Combining with representation learning [126] Local model LW-PLS [71,72,73,74] JITL-Based KPLS [75,76,77,78] Combining with other modeling strategies [79,80,81,82] Adaptive SVM [84,85] JITL based SVR [87,88,89,34,90] Combining with outlier detection [86,91] Combining with other modeling strategies [92,93] JITL based RVM [94,95] Dynamic RVM [96] Semi-supervised RVM [97,98] Combining with other modeling strategies…”

A deep‐red xanthene‐based highly sensitive fluorescent probe for detection of hypochlorite

“…The low cytotoxicity and high solubility in aqueous medium of probe TPE‐SQ5 enabled it to be suitable for bioimaging applications. Wang et al [15] synthesized near‐infrared (NIR) fluorescent probe PyOX for detecting the ClO − . Due to the oxidation properties of HClO, the recognition group aldoxime was converted to nitrile group when the PyOX was exposed to HClO and exhibited strong turn‐on red fluorescence centered with 680 nm.…”

Gaussian mixture model sample selection strategy–based active semi-supervised soft sensor for industrial processes

“…In the actual industrial processes, there are many obstacles, such as harsh environments and significant data acquisition costs, to real-time detection of key process parameters. Soft sensor technology can help addressing this problem by estimating the difficult-to-measure variables using other easy-to-measure process variables (Wang et al, 2021; Yan et al, 2020). In recent years, with the advances in data acquisition and storage technologies and data analysis methods, the data-driven soft sensor modeling methods, especially deep neural networks, have received considerable attention in process modeling (Liu et al, 2018b; Yuan et al, 2019).…”