IDENTIFICATION OF FLUID SYSTEMS USING GENETIC PROGRAMMING

In recent years, applied researchers have become increasingly interested in Adaptive Search (AS), techniques such as the Genetic Algorithm (GA) [1], and Genetic Programming GP [2], for engineering design [3]. This paper illustrates the effectiveness of the genetic programming paradigm for simple fluid systems identification problems. The objective of the paper is to establish methods for systems identification using GP and sets of empirical data. The manipulation and optimisation of these approximate functions that describe the physical process is achieved using the GP approach and by the development of complementary AS techniques. Two new GP operators are introduced, the first searches through possible values of terminals for a particular functional tree structure, and the second uses functional induction [4] to improve the performance of the technique