Genetic Algorithms for dynamic land-use optimization

Jin, Nanlin, Termansen, Mette and Hubacek, Klaus (2008) Genetic Algorithms for dynamic land-use optimization. In: 2008 IEEE Congress on Evolutionary Computation (IEEE World Congress on Computational Intelligence). IEEE, Piscataway, NJ, pp. 3816-3821. ISBN 978-1-4244-1822-0

Full text not available from this repository. (Request a copy)

Abstract

This paper concerns the use of Genetic Algorithms designed to optimize agricultural land use based on economic criteria. The agricultural areas considered are heather moorland areas in the UK where sheep farming competes with grouse farming and the land is managed differently for each activity. Additionally, there are tenant farmers who rent land for fixed periods and are more interested in short term economic gain and landlords who are more concerned with land value and capability and economic returns in the longer term. This paper explores the application of Genetic Algorithms (GAs) to what we call an inter-temporal optimization. Inter-temporal optimization aims to maximize outcomes for a period of time, not for a time point. GAs are shown to be able to cope with two important features of intertemporal optimization: (1) dynamics; (2) optimizing areas of landscape. These two features make it difficult for traditional approaches such as econometrics and mathematical dynamic programming to tackle such an optimization problem. This paper exemplifies GA's capabilities by tackling an intertemporal optimization problem in land-use decision making. We use GA to represent land-use decisions, to simulate economic and biologic dynamics, and to optimize decisionmakers' objectives in inter-temporal optimization. Experimental results indicate that a long-term inter-temporal optimization smoothes the impacts of dynamics and reduces the number of decision changes. We also compare the experimental results versus the predictions made by agricultural experts. We have found that a GA system forecasts land-use changes in line with experts' predictions. This work demonstrates how GA successfully deals with dynamics for inter-temporal optimization.

Item Type:	Book Section
Subjects:	G400 Computer Science
Department:	Faculties > Engineering and Environment > Computer and Information Sciences
Depositing User:	Becky Skoyles
Date Deposited:	11 Aug 2014 10:45
Last Modified:	12 Oct 2019 22:29
URI:	http://nrl.northumbria.ac.uk/id/eprint/17373

Downloads