Successful Application of Model Based Predictive Control for Production and Thermal Efficiency Optimization of High Temperature Melters

- Organization:
- The Minerals, Metals and Materials Society
- Pages:
- 7
- File Size:
- 994 KB
- Publication Date:
- Jan 1, 2012
Abstract
"In the past decade, advanced multiple input/multiple output (MIMO) process control systems have found their way into high temperature melting processes. Today a growing number of high temperature melters have been equipped with a supervisory control system. Daily regulation of fossil fuel firing and electric energy supply to stabilize temperatures is no longer in the hands of the operator, but fully taken over automatically by the ES III™ MPC controller, which provides consistent process control, 24 hours per day, focused to operate the entire glass production process in the most efficient way. The presentation will show practical results and benefits achieved in the high temperature glass and cement industry.IntroductionThe Expert System, either version ESII™ or ESIII™, provides supervisory multi-variable model based predictive control in time and space This definition is not very understandable to people who are not familiar with advanced process control, thus let's try to explain it word by word.The word supervisory means that the Expert System implementation preserves the primary control level: all devices (PLCs, internal switches) as well as up to date control mechanisms (PID loops, availability to change inputs manually). ES III™ can even use primary control level elements as part of the advanced process control.The next word, multi-variable, means that multiple inputs can be handled at the same time to affect multiple outputs. This property is very important for supervisory control of the glass melting and conditioning processes. For example, in a float furnace the total amount of gas is only one of the inputs that affect the crown temperatures. The next one is batch feed rate. And batch composition (the next one) is maybe even more important. An ambient temperature also correlates with crown temperatures. So it can be concluded that crown temperatures are affected not only by one variable ""total gas"", but by many more. On the other hand, there are usually several crown temperatures with different importance for process control. And we have not discussed yet the effects of glass level, canal temperatures, oxy-boosting, etc. All of them affect (or are affected by) process control too. Overall, the process control is a complex problem where multiple inputs and outputs play strongly together at various levels of importance."
Citation
APA:
(2012) Successful Application of Model Based Predictive Control for Production and Thermal Efficiency Optimization of High Temperature MeltersMLA: Successful Application of Model Based Predictive Control for Production and Thermal Efficiency Optimization of High Temperature Melters. The Minerals, Metals and Materials Society, 2012.