Knowledge driven optimisation has been developed in an attempt to overcome difficulties in applying existing reactor network synthesis methods to complex systems. Knowledge derived from kinetic relationships is applied to superstructure optimisation in the form of a customised rule-based Tabu Search where rules are used to guide optimisation decisions. Nonisothermal behaviour is represented using temperature profiles along the length of a reactor. Results show the method outperforms a standard Tabu Search both in solution quality and speed of convergence.
- Knowledge driven optimisation
- data mining
ASJC Scopus subject areas
- Chemical Engineering(all)
- Computer Science Applications