Developing a Soft Sensor for MTBE Process Based on a Small Sample
- PDF / 643,913 Bytes
- 11 Pages / 612 x 792 pts (letter) Page_size
- 69 Downloads / 154 Views
TOMATION IN INDUSTRY
Developing a Soft Sensor for MTBE Process Based on a Small Sample S. A. Samotylova∗,a and A. Yu. Torgashov∗,b ∗
Institute of Automation and Control Processes, Far Eastern Branch of the Russian Academy of Sciences, Vladivostok, Russia e-mail: a [email protected], b [email protected] Received December 15, 2018 Revised December 15, 2018 Accepted July 9, 2020
Abstract—The paper discusses the development of a soft sensor (SS) for an MTBE plant in the cases when the training sample either is small or does not comprise the whole of quality range because of process plant non-stationarity as well as the difficulty and high cost of retrieving more information. In order to build a soft sensor that provides higher accuracy in estimating the quality of the output product, an algorithm for extending the initial training sample using a rigorous model of the distillation column of the methyl-tert-butyl ether production under condition of exactly unknown values of the Murphree trays efficiency has been proposed. The resulting SS enables 40% improvement of product quality prediction. Keywords: predictive modeling, soft sensor, distillation column, extended training sample DOI: 10.1134/S0005117920110120
1. INTRODUCTION One of the main problems facing the oil refining and petrochemical industry is improving the quality of the main types of oil products and their economic efficiency. The achievement of these goals is possible not only through the modernization of the oil refineries themselves, but also process control systems. The results of the analyses obtained by the means of factory laboratories, as a rule, do not have the necessary completeness and speed, since the measurements are carried out once or twice a day, which forces technologists to maintain modes that provide a large margin for the quality of products, thereby increasing the consumption of raw materials and energy. The use of on-stream analyzers significantly increases the efficiency of material flow monitoring, however, they have high cost and require constant calibration. This problem can be solved by using soft sensor (SS), which allow you to quickly track changes in the quality of product flows. Implementation of SS ensures prompt quality control of output products with minimal energy consumption and losses. SS can be integrated into advanced process control (APC) systems. Scientific groups of the Institute of Control Sciences of the Russian Academy of Sciences had a great influence on the development of APC systems in industry [1]. In the 1960s, the laboratory headed by Professor N.S. Raibman was one of the first in the world to deal with identification issues—the construction of mathematical models of real objects, processes and systems according to experimental data. The main theoretical achievements of the Laboratory include: —justification and development of dispersion methods for identification of nonlinear objects; —development of the theory of adaptive identification of non-stationary systems; —setting and justification of the minimax (game) app
Data Loading...
