Title: A grey-encoded hybrid accelerating genetic algorithm for parameter optimisation of environmental models

Authors: Xiaohua Yang, Zhifeng Yang, Zhenyao Shen, Jianqiang Li

Addresses: School of Environment, State Key Laboratory of Water Environment Simulation, Beijing Normal University, Beijing 100875, China. ' School of Environment, State Key Laboratory of Water Environment Simulation, Beijing Normal University, Beijing 100875, China. ' School of Environment, State Key Laboratory of Water Environment Simulation, Beijing Normal University, Beijing 100875, China. ' Water Strategy Research Division, Water Resources and Hydropower Planning and Design General Institute, Beijing 100011, China

Abstract: A new method, grey-encoded hybrid accelerating genetic algorithm, is presented for the parameter optimisation of environmental models. With the shrinking of searching range, the method gradually directs to optimal result with the excellent individuals obtained by grey genetic algorithm embedding Nelder-Mead simplex searching operator. The convergence theorem is given for guaranteeing the global convergence of the new genetic algorithm. The global optimisation of the new genetic algorithm is analysed. Its efficiency is verified by application of ten test functions. The comparison of our algorithm with six other algorithms is presented. This algorithm overcomes any Hamming cliff phenomena in existing genetic algorithms, and it is good for the parameter optimisation for the practical O|Connor water quality model and rainfall-runoff model.

Keywords: environmental models; global optimisation; grey-encoded genetic algorithms; Nelder-Mead simplex algorithm; accelerating convergence; modelling; environmental pollution; water quality; rainfall runoff.

DOI: 10.1504/IJEP.2006.011209

International Journal of Environment and Pollution, 2006 Vol.28 No.3/4, pp.239 - 252

Published online: 06 Nov 2006 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article