Automatic tuning for process control
In polymer industries, mixing of chemicals are very common. When two or more chemicals are mixed together, the mixer produces a sudden unpredictable and unmeasurable heat due to a nonlinear exothermal reaction. Resin adhesives are produced through a mixing process between phenol and formaldehyde wit...
Saved in:
| Main Author: | |
|---|---|
| Format: | Research Report |
| Language: | English English |
| Published: |
1998
|
| Subjects: | |
| Online Access: | https://eprints.ums.edu.my/id/eprint/30813/5/ABSTRACT.pdf https://eprints.ums.edu.my/id/eprint/30813/2/Automatic%20Tuning%20For%20Process%20Control.pdf https://eprints.ums.edu.my/id/eprint/30813/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In polymer industries, mixing of chemicals are very common. When two or more chemicals are mixed together, the mixer produces a sudden unpredictable and unmeasurable heat due to a nonlinear exothermal reaction. Resin adhesives are produced through a mixing process between phenol and formaldehyde with certain catalyst. This mixture has to undergo a specific heating process to achieve the required quality of resin. Developing a suitable methodology to control the exothermal heat in producing resin adhesives is the main task of this research effort. Based on simulation studies in MATLAB-SIMULINK and real time implementations of the process control, the predictive loop in FLC structure has been proved to be very effective in predicting the future temperature error and change in error and computing the inputs needed for FLC to decide the switch ON duration for heating or cooling actions. Thus, PFL controller can be used to compensate the inherent time delay of the reactor system. However, as shown in the simulation and real time results, the steady state performance of the process is affected by the slowly varying un-modelled parameters. Therefore, in order to reduce the effect of un-modelled parameters on steady state errors in temperature responses, an adaptive control approach is applied to the PFL control structure as an outer loop. This has led to the development of the APFL controller for further improving the dynamic performance of the exothermic process. The suitability of this complex, but very effective APFL controller is thus ascertained through simulated experiments. Based on the performance studies, it can be concluded that the APFL is capable of overcoming the shortfalls due to the inherent time delay and steady state error.
Pl D control is one of the general solutions often applied in the chemical production line.- Due to the inherent time delay of the industrial plant, the experienced operators have to tune the PIO gains using a trial and error method. However, this method has been found to be a source of errors that affects the quality of the product. In addition, the PIO controller needs a precise mathematical model for on-line temperature control. This has become a problem, as the nonlinear exothermal process is not easy to be modelled due to the plant parameters varying unpredictably. The main objective of this thesis is to propose a model-free control system, which is precise in temperature control. A computer based temperature control scheme has been developed in this work, which employs fuzzy logic control (FLC) principles. Two types of control methods are suggested, namely predictive fuzzy logic (PFL) and adaptive predictive fuzzy logic (APFL). The developed real time PFL controller has been applied to a specially fabricated pilot process plant for testing its accuracy in controlling and regulating the reactant temperature. The pilot plant has facilities for heating and cooling the mixture. |
|---|