Articles

Ž. Ujević Andrijić, I. Mohler, N. Bolf, H. Dorić

Industrial facilities show an increasing need for continuous measurement and monitoring a large number of process variables due to strict product quality requirements, environmental laws and for advanced process control application. On-line analysers typically suffer from long measurement delays not desirable in continuous control. Suitable alternative are soft sensors and inferential control. In this paper the development of soft sensor models for the estimation of light reformate benzene content is carried out. Linear dynamical autoregressive model with external inputs (ARX), autoregressive moving average model with exogenous inputs (ARMAX) and Box-Jenkins (BJ) models are developed. For the regression vector optimization usually performed by trial and error, Genetic Algorithm (GA) and Simulated Annealing (SA) methods have been applied. The results indicate that the GA and SA as global optimization methods are suitable for the regressor order estimation of linear dynamical models with multiple inputs. Based on developed soft sensors, it is possible to apply advanced process control schemes.

For more information please send us an email.

Mohler, N. Bolf, G. Galinec, M. Golob

This paper elaborates methods of soft sensor development for dynamic model identification and process control of Sulphur Recovery Unit (SRU) in refinery production. Experimental data are acquired from refinery unit and include available on-line measured variables and on-line analysis. The results are soft sensor models for optimal control of SRU with aim to minimize SO2 and H2S emissions. The soft sensors were developed conducting multiple linear regression analysis and using neural network-based and fuzzy logic models. From a variety of different model structures the best results were achieved with multi-layer perceptron’s and neuro-fuzzy soft sensor models.

For more information please send us an email.

Rogina, I. Šiško, I. Mohler, Ž. Ujević, N. Bolf

Due to the strict norm requirements of keeping products in crude refining units within specifications, laboratory testing and quality control of the products are necessary. Given this reason, virtual soft sensor for continuous quality estimation of light naphtha as the crude distillation unit (CDU) product was developed. Experimental data included available continuous measurements of CDU process streams (temperatures, pressures and flowrate) and laboratory analyses undertaken twice a day. The results are soft sensor models for light naphtha vapor pressure (RVP) estimation. Soft sensor models have been developed conducting multiple linear regression analysis and using neural network-based models such as LNN, MLP and RBF. Considering statistical and sensitivity analysis, the best results for both oils were obtained with MLP and RBF neural networks. The results show possible application of the soft sensor models for estimating light naphtha RVP as an alternative for laboratory testing

For more information please send us an email.

Novak, I. Mohler, M. Golob, Ž. Ujević Andrijić, N. Bolf

Due to growing fuel quality demands, continuous measurements of process variables and product quality properties in the crude distillation unit (CDU) are necessary. One of the key diesel fuel properties is kerosene cold filter plugging point (CFPP). CFPP is usually determined only by laboratory assays. On the basis of available continuous measurements of temperatures and flows of appropriate process streams, soft sensor models for the estimation of kerosene CFPP have been developed. Data pre-processing includes: detection and outlier removal, generating additional output data by Multivariate Adaptive Regression Splines (MAR Splines) algorithm, detrending data and filtering data. Soft sensors are developed using linear and nonlinear identification methods. Model structures are optimized by Genetic Algorithm (GA) and ANFIS (Adaptive Neuro-Fuzzy Inference System) algorithm. Results of the Output Error (OE) model, Hammerstein–Wiener (HW) model and neuro-fuzzy model are shown. Developed models were evaluated based on the final prediction error (FPE), root mean square error (RMSE), mean absolute error (AE) and FIT values. The best results are achieved with neuro-fuzzy model.

For more information please send us an email.

Bolf, I. Mohler
In today’s globally competitive marketplace, industrial plants are looking at new ways to increase plant efficiency, production rates, safety and reliability. Engineer education and training, monitoring, diagnosis and plant optimization play a key role in satisfying technological, economic and environmental constraints. Furthermore, control system optimization is the basis for system improvement and advanced process control (APC) implementation. Very few plants use modern software for control quality monitoring, controller tuning, APC or optimization. The reasons are absence of engineering knowledge and unavailability of practical and robust process control software tools for system identification, parameter optimization and control quality monitoring, running plants conservatively due to fear of causing shutdowns and plant problems. Process control software tools for quick and easy system identification using available data from the plant’s historian can help tremendously improve the control quality and the plant’s profit margin. It is possible to analyze multivariable systems, complex, nonlinear and slow processes with long dead times and long-time constants commonly encountered in process industry. Optimization of primary and advanced control schemes stabilizes the process and allows the plant to run closer to process, equipment and economic constraints. This increases production rates, minimizes operating costs and improves product quality. The overall control system performance is significantly improved which ultimately has a positive effect on product quality and energy consumption thus proving application of control system diagnostics and optimization usefulness.

For more information please send us an email.

Mohler, Ž. Ujević Andrijić & N. Bolf

Industrial facilities nowadays show an increasing need for continuous measurements, monitoring and controlling many process variables. The on-line process analysers, being the key indicators of process and product quality, are often unavailable or malfunction. This paper describes development of soft sensor models based on the real plant data that could replace an on-line analyzer when it is unavailable, or to monitor and diagnose an analyser’s performance. Soft sensors for continuous toluene content estimation based on the real aromatic plant data are developed. The autoregressive model with exogenous inputs, output error, the nonlinear autoregressive model consisted of exogenous inputs and Hammerstein–Wiener models were developed. In case of complex real-plant processes a large number of models regressors and coefficients need to be optimized. To overcome an exhaustive trial-and-error procedure of optimal model regressor order determination, differential evolution optimization method is applied. In general, the proposed approach could be, of interest for the development of dynamic polynomial identification models. The performance of the models is validated on the real-plant data.

For more information please send us an email.

Mohler, G. Galinec, N. Hölbling, N. Bolf, Ž. Ujević

Virtual soft sensors are developed for properties estimation of diesel fuel as the crude oil column side product. Because of the growing standards for the fuel quality and needs to produce various gradations of diesel fuel, frequently laboratory testing and quality controls of the products are necessary. On the basis of available continuous temperature measuring of particular process streams, soft sensors for estimating end boiling point (D95) of diesel fuel have been developed. Linear and nonlinear soft sensor models have been built using linear regression and artificial neural networks. Statistical data analysis has been carried out and the results were critically judged.

For more information please send us an email.

Howes*, I. Mohler**, N. Bolf

In today’s globally competitive marketplace, industrial plants are looking at new ways to increase plant efficiency, production rates, safety and reliability. Engineer education and training and plant optimization play a key role in satisfying technological, economic and environmental constraints. Furthermore, control system optimization is the basis for system improvement and advanced process control (APC) implementation. Very few plants use modern software for controller tuning and simulation, APC or optimization. The reasons are absence of engineering knowledge and unavailability of practical and robust process control software tools for system identification, simulation and parameter optimization, running plants conservatively due to fear of causing shutdowns and plant problems. Presented process control software tools for process simulation and tuning education for quick and easy multivariable closed-loop system identification using available data from the plant’s historian can help tremendously improve control education and plant operation.

For more information please send us an email.