Inferential Active Disturbance Rejection Control of a Distillation Column

Kalbani, Fahad Al; Zhang, Jie

doi:10.1016/j.ifacol.2015.09.001

Cited by 5 publications

(2 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It should be noted that for highly nonlinear processes, such as batch distillation processes, nonlinear soft sensors should be utilized [ 26 , 27 , 28 , 29 ]. In our earlier work [ 30 ], static PCR models are used. In the current work, the soft sensors are extended to using dynamic PCR models.…”

Section: Inferential Adrc Schemementioning

confidence: 99%

“…Therefore, six principal components are used in the top composition model and 10 principal components are used in the bottom composition model. The developed PCR models for top and bottom product compositions are as follows [ 30 ]:

where y D and y B represent the top and bottom compositions (wt%) respectively, and Δ T is the deviation of a tray temperature from its nominal mean value.…”

Section: Pcr Model-based Software Sensorsmentioning

confidence: 99%

See 1 more Smart Citation

Inferential Composition Control of a Distillation Column Using Active Disturbance Rejection Control with Soft Sensors

Kalbani¹,

Zhang²

2023

Sensors

View full text Add to dashboard Cite

This paper presents the integration of active disturbance rejection control (ADRC) with soft sensors for enhancing the composition control performance in a distillation column. Static and dynamic soft sensors are developed to estimate the top and bottom product compositions using multiple tray temperatures. In order to cope with the collinearity issues in tray temperature measurements, static and dynamic principal component regression is used in developing the soft sensors. The soft sensor outputs are introduced as the feedback signals to ADRC. This control scheme is termed as “inferential ADRC control”. Static control offsets are eliminated through mean updating in the soft-sensor models. The effectiveness of the proposed control scheme is demonstrated on a benchmark simulated methanol-water distillation column.

show abstract

Section: Inferential Adrc Schemementioning

confidence: 99%

where y D and y B represent the top and bottom compositions (wt%) respectively, and Δ T is the deviation of a tray temperature from its nominal mean value.…”

Section: Pcr Model-based Software Sensorsmentioning

confidence: 99%