Fuzzy Optimization Approach for the Synthesis of Polyesters and Their Nanocomposites in In-Situ Polycondensation Reactors

A proportional-derivative (PD) fuzzy controller was presented to control the temperature of a polycondensation reactor. Synthesis of various polyesters and copolyesters were conducted using the devised controller. This controller uses reactor temperature error as well as its derivative. The desired jacket temperature had an uncertainty that was affected by noise and disturbance. This uncertainty was modeled using fuzzy numbers, and a fuzzy trajectory was achieved for the desired reactor temperature. Then, a generalized Takagi-Sugeno (GTS) fuzzy controller was designed. An adaptation algorithm was included in the controller. Compared to that of conventional proportional-integral-derivative (PID) controllers, experimental results presented a very good performance to control the reactor temperature of in-situ polymerization of polyesters. Although the pm controller finally controlled the temperature with an accuracy of +/- 6 degrees C, the temperature overshoot was around 20 degrees C which could completely degrade the polymer. However, the fuzzy PD controller had no overshoot. The performance of the fuzzy PD controller to control a suitable temperature of the reactor was excellent with 0.2% accuracy without overshoot. The NMR results confirmed that the final polymers have-been synthesized successfully. Moreover, the molecular weights of the samples, synthesized under fuzzy control, were generally higher and there was a better intercalation structure in such samples.