Knowledge-Based System for Forecasting Snow Avalanches of Chowkibal-Tangdhar Axis (J&K
Abstract
In this paper an attempt has been made to use artificial intelligence in avalanche forecasting and to develop arule-based expert system for predicting direct action avalanches of Chowkibal-Tangdhar axis (JLK). Using C-language integrated production system (CLIPS), procedural knowledge is represented in the form of rules. Thecondition attributes of the rulebased system are 28 variables selected from 1 154 samples of snow-met and snowprofile data.The relative contribution of each variable on avalanche days and non-avalanche days and their influence on sitewise release of avalanche was studied to formulate 358 rules. These rules, which include 173decision rules, were finally implemented and validated for running the model. Sixty-three samples of snow-metdata and pit profile data attributing to avalanche days and 54 samples of non-avalanche days were run on themodel. The results show that the knowledge-based model can predict avalanche days with 76 per cent efficiency.The misclassified results accounted for 28.2 percent of 117 test samples.References
McClung, D.M. & Schaerer, P.A. The avalanche handbook. Seattle, WA, The Mountaineers.
La Chappelle, E.R. The fundamental processes in conventional avalanche forecasting. Journal of Glaciology, 1980,26(94), 75-84.
Ganju, A.; Prashar, D.K. & Thakur, N. Analysis of process-oriented avalanche forecasting
technique followed at SASE, Manali, India. Proceedings of the International Snow Science
Workshop, 27 September -01 October 1998. Bill Williamson, Stevens Pass Ski Area, P.O. Box 98, Skykomish, WA 98288,1998. pp. 419-25.
McClung, D.M. & Tweedy, J. Numerical avalanche prediction: Kooteney pass, British
Columbia, Canada. Journal of Glaciology, 1994, 40(135), 350-58.
Buser, 0.; Butler, M. & Good, W. Avalanche forecast by nearest neighbour method.
International Association of Hydrological Sciences Publication No- 162. pp. 557-69.
Schweizer, J. & Fohn, M.B. Paul. Avalanche forecasting: an expert system approach. Journal of Glaciology, 1996,42(141), 318-40.
Naresh, P. & Pande, B. Numerical avalanche prediction by statistical and AI techniques.
Proceedings of the Workshop-97, 24-25 October 1997. Snow & Avalanche Study Establishment, Manali, India, 1997.
Forgy, C. Rete. A fast algorithm for the many pattern match. object pattern match problem. Artificial Intelligence, 1982,19, 17-37.
Where otherwise noted, the Articles on this site are licensed under Creative Commons License: CC Attribution-Noncommercial-No Derivative Works 2.5 India