Modeling of air temperature using artificial intelligence in various climates of Iran

Amini Rakan, A. (2013) Modeling of air temperature using artificial intelligence in various climates of Iran. Masters thesis, University of Zabol.

[img]
Preview
Text
Modeling of air temperature using artificial.pdf

Download (75kB) | Preview

Abstract

Air temperature is an important meteorological and climatological factor that its changes are sources of the most fluctuations on agriculture, water resources and environment. The measurement and prediction of air temperature are of high importance and have a longer background than the other atmospheric parameters. The aim of this study is to illustrate a mathematical model for the time series of monthly mean temperature, using a new approach of gene expression programing (GEP). (GEP) is a powerful tool in order to model and detect the linkage between complex phenomena that produced from Genetic programming. Also another characteristic of this approach is to present mathematical relationship from the produced model. This approach is performed in two main steps: 1. train the model that can predict time series. 2. Test or validate the model that produced in last step with actual data. Models are instructed in six different historical (subsequence) patterns. In the second step for validating models, Root Mean Square Error (RMSE) and coefficient of determination (R2) are used. 80 percent of data is considered for training and 20 percent considered for model verification. In this way time series of monthly mean temperature from 31 synoptic stations of Iran is modeled. Models classified in seven different climate of Iran. Reasons of each climate were presented and were compared together. Results showed that 45 percent of models produced in fourth pattern. Genetic programing is a very suitable approach for modeling hydrological parameters, although its performance is different in various climates. For models that are produced from specific climate, it can be expected that, main factors like input patterns are similar. Finally results showed that the (GEP) for modeling of monthly mean temperature is very suitable. The best model in this research was achieved for seventh climate group and Zabol city.

Item Type: Thesis (Masters)
Subjects: T Technology > TA Engineering (General). Civil engineering (General)
Depositing User: admin admin1 admin2
Date Deposited: 25 Oct 2016 05:21
Last Modified: 25 Oct 2016 05:21
URI: http://eprints.uoz.ac.ir/id/eprint/887

Actions (login required)

View Item View Item