Fuzzy Rule Generation from the EMEP Ozone Model to Examine Source-Receptor Relations

Ryoke, M. (1996). Fuzzy Rule Generation from the EMEP Ozone Model to Examine Source-Receptor Relations. IIASA Working Paper. IIASA, Laxenburg, Austria: WP-96-130

Preview

Text WP-96-130.pdf Download (551kB) | Preview

Abstract

The objective of this paper is to describe research on the development of a simplified version of the European ozone model using fuzzy rule generation methodology. The ozone model is used to predict tropospheric (at the ground level) ozone concentration. The simplified ozone model illustrates source-receptor relationships between ozone precursor emissions (NOx and VOCs) and ozone concentration in the troposphere, taking into account meteorological conditions. This ozone model was developed by the Cooperative Programme for Monitoring and Evaluation of Long-Range Air Pollutants in Europe (EMEP). The EMEP model provides a detailed prediction of ozone concentration at every grid in Europe by taking into account physical and chemical mechanisms. However, the model is too complicated for nonspecialists, such as policymakers trying to set emission reduction levels that result in ozone concentrations below given limits. Therefore, there is a need for a simplified ozone model that can be verified by the EMEP model and that can be used for analyzing policy options.

One approach is to use the fuzzy rule generation methodology. In this approach, the simplified model consists of a number of fuzzy rules. Fuzzy rules have a fuzzy proposition in the conditional statement and a linear regression model in the conclusion. The rules describe a complete nonlinear system by using several linear models and membership functions. The development of such fuzzy rules is called fuzzy modeling. The membership functions of conditional variables are determined by the subset of data which is obtained by a clustering method. The degree of confidence of a rule is determined by the grade of the membership functions for input values. The role of fuzzy logic is to integrate fuzzy rules smoothly.

In this paper, a basic scenario, which predicts no reduction of ozone precursor emissions, is used to determine fuzzy rules, subsequent scenarios are derived from the basic scenario, which includes information on source-receptor relationships. Simplified models of three grids have been developed to show the effectiveness of this approach. This methodology can be used to develop models of all grids.

Actions (login required)

View Item