City Research Online

Abstract

This research investigates on-ine steady-state optimisation techniques based on the augmented Lagrangian function. The augmented Lagrange function is applied to a class of hierarchical Integrated System Optimisation and Parameter Estimation (ISOPE) algorithms. The ISOPE algorithm employs model equations and information feedback from the process. In order to overcome the model uncertainty, the algorithm needs perform both optimisation and identification tasks. The identification task is to iteratively update the model parameter value which will be used to improve the modified optimisation problem.

Two different methods of augmentation are developed for augmenting the ISOPE algorithm: the variable and constraint augmentation. The applicability of these new versions of augmented algorithms are extended to non-convex problems. The constraint augmented algorithms with double iterative coordination strategies are found to be the most efficient algorithms in reducing the number of required set-point adjustments compared to the other algorithms which have been proposed.

The optimality and convergence properties of these augmented algorithms are derived. Bsed on the convergence studies and simulation results, a comparative study on the applicability of the algorithms aer performed.

Among the proposed augmented ISOPE algorithms, the model based double loop algorithm is the most suitable for on-line application and further work can be performed to investigate the on-line behaviour of the algorithm. The author also hopes that this work will provide valuable information for further development of the ISOPE algorithms.

Publication Type:	Thesis (Doctoral)
Subjects:	T Technology > T Technology (General) T Technology > TA Engineering (General). Civil engineering (General)
Departments:	School of Science & Technology > Engineering School of Science & Technology > School of Science & Technology Doctoral Theses Doctoral Theses

Export

Downloads