Time series data-based modeling and real-time optimization for the temporal feature extraction of lube base oil plant

Abstract

In lube base oil production, a vacuum distillation unit is a crucial unit operation which must be controlled tightly to meet final product specifications. However, it is hard to analyze the process accurately due to the instability of the online measurement and there is no rigorous protocol for the process operation. To address this problem, model-based optimal control can be employed, which requires an accurate model relating to the operation and quality variables involved. However, the mathematical modeling of VDU is hampered by the lack of process knowledge and the limited online data acquisition capability for key quality variables. As a recourse, data-based modeling can be attempted to develop inferential sensors and controllers but one is faced with the problem of temporal credit assignment due to the complex characteristics of the VDU process dynamics, e.g., varying time-lags among the variables. To overcome this difficulty, the use of a stacked RNN structure and attention mechanism with sequence length are proposed for the prediction of kinematic viscosity. In addition, regression models for distillation temperature are constructed through the investigation of a huge model structure library. Then, these quality predictors are applied in developing a real-time optimizer which adjusts key operating variables in order to satisfy the product specifications despite various disturbances. Accuracies of the soft sensor and the optimization model are evaluated, first offline and then, online through a commercial plant test performed at SK Ulsan Refinery Complex in Korea. The overall approach and framework are general and can be used for the development of inferential sensors and controllers in other similar situations.