Economic modelling with a general purpose GA

This paper describes the design and implementation of a general-purpose Genetic Algorithm (GA) in object-oriented C + +. The GA features include the concepts of DNA and phenocell where DNA contains genetic coding for the phenocell. Intelligent features of the GA include life span and parasites. Life span indicators limit explosion of the population size by giving fitter members a longer life span. Parasitic operators are used to incorporate varying degrees of randomness in the system. Descriptions of how the GA solves an optimization problem such as the travelling salesman problem (TSP) are given. The GA is then used to model pricing behaviours such as those of the Australian Petroleum Industry (API). In contrast to the TSP, a 2-Population model is needed to show realistic interactions between Companies and Customers. Initial implementation of this research suggests that the design of the GA is easily adoptable to solve problems requiring more than 1 population. Current work in this research involves the use of actual data.