City Research Online

Abstract

This research is concerned with the problem of determining optimum operating conditions of a system where a mathematical model is employed to compute the performance index. A family of algorithms known as Integrated System Optimisation and Parameter Estimation (ISOPE) algorithms have been developed. These algorithms are based on the Modified Two-step approach, where the model contains certain free parameters which are estimated periodically using system measurements, and are then used to solve an optimisation problem in an iterative manner. An advantage with these methods is that the optimum operating conditions are obtained even when the model structure is not identical to that of the system. It is then quite evident that the model-reality differences can be described by the magnitude of the parameters which are estimated. However, such a simplistic notion of model-reality differences is not an accurate description of these differences. For often it is possible that the value of the parameters estimated are small, but the efficiency of the iterative process could be poor. Then one of the main aim of this research is to identify what constitutes these model-reality differences, and how do the ISOPE algorithms overcome these differences.

The ISOPE algorithms studied in this research cover both the centralised as well as the decentralised cases. However, a detailed analysis of the model-reality differences, is under taken for the centralised case. For it can be seen that the analysis is also valid for the decentralised cases.

The ISOPE algorithms have many similarities. They are derived in a similar manner, the algorithmic and control structures are similar, so also is the problem formulation for the various algorithms. It is in this context that a generalisation of these algorithms is attempted.

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

Export

Downloads