Abstract:Soft sensors have been widely accepted for online estimating key quality-related variables in industrial processes. The Gaussian mixture models (GMM) is one of the most popular soft sensing methods for the non-Gaussian industrial processes. However, in industrial applications, the quantity of samples with known labels is usually quite limited because of the technical limitations or economical reasons. Traditional GMM-based soft sensor models solely depending on labeled samples may easily suffer from singular covariances, overfitting, and difficulties in model selection, which results in the performance deterioration. To tackle these issues, we propose a semisupervised Bayesian GMM (S(2)BGMM). In the S(2)BGMM, we first propose a semisupervised fully Bayesian model, which enables learning from both the labeled and unlabeled datasets for remedying the deficiency of infrequent labeled samples. Subsequently, a general framework of weighted variational inference is developed to train the S(2)BGMM, such that the rate of learning from unlabeled samples can be controlled by penalizing the unlabeled dataset. Case studies are carried out to evaluate the performance of the S(2)BGMM through a numerical example and two real-world industrial processes, which demonstrate the effectiveness and reliability of the proposed approach.

A Secure Estimator with Gaussian Bernoulli Mixture Model

Secure Estimation And Event-Triggered Control For Cyber-Physical Systems Based On Intermediate Estimator

Secure State Estimation Against Integrity Attacks: A Gaussian Mixture Model Approach

Secure State Estimation Based on Gaussian Mixture Model for Systems under Extra Packet Injection Attacks

Event-triggered Attack Detection and State Estimation Based on Gaussian Mixture Model

Event-triggered State Estimation for Cyber-Physical Systems with Partially Observed Injection Attacks

An Adaptive Gaussian Process‐based Method for Efficient Bayesian Experimental Design in Groundwater Contaminant Source Identification Problems

Secure State Estimation for Linear Time-varying Processes Via Local Estimators

Security Analysis of Continuous-Time Cyber-Physical System Against Sensor Attacks

Short-term Traffic Safety Forecasting Using Gaussian Mixture Model and Kalman Filter

Semisupervised Bayesian Gaussian Mixture Models for Non-Gaussian Soft Sensor

Generalized Pseudo-Bayesian Estimator for Networked Control Systems over UDP-like Channels.

Security against false data injection attack in cyber-physical systems

Robust Cubature Kalman Filter With Gaussian-Multivariate Laplacian Mixture Distribution and Partial Variational Bayesian Method

On Finite-State Stochastic Modeling and Secure Estimation of Cyber-Physical Systems.

State Estimation in Unknown Non-Gaussian Measurement Noise Using Variational Bayesian Technique.

A Bayesian Filtering Algorithm for Gaussian Mixture Models

Approximate MMSE Estimator for Linear Dynamic Systems With Gaussian Mixture Noise

Secure Dynamic State Estimation by Decomposing Kalman Filter

An Interacting Multiple Model Approach for State Estimation with Non-Gaussian Noise Using A Variational Bayesian Method

Secure Dynamic State Estimation Via Local Estimators