Multirate Autoregressive Dynamic Latent Variable Model for Soft Sensing in Dynamic Processes

Using data-driven technology for soft sensing can significantly reduce the difficulty of obtaining key quality variables in industrial processes. Under real industrial conditions with diverse sensors, the measured data inevitably contain complex multirate and dynamic characteristics. To improve accuracy in such multirate dynamic processes, a multirate autoregressive dynamic latent variable (Mr-ARDLV) model is proposed for dynamic soft sensing. The model is built using the expectation-maximization (EM) algorithm, introducing two sampling indicators to extract high-order autocorrelation and cross correlation between multirate variables. The model’s effectiveness has been preliminarily validated through a numerical case, and it has also demonstrated excellent performance in two real cases: a petrochemical process and a papermaking wastewater treatment process.