Time Series Analysis Based On The Interval Cloud Model

At present, the interval data widely exists in our daily life and industrial production. The analysis and prediction about the interval data gets more attention and more importance as well. This paper introduces the cloud model. The cloud model is an uncertain transformational model. It achieves an uncertain transformational model from quantitative value to qualitative concept and associates randomness and fuzziness in the process. The cloud theory includes the cloud model, the cloud transformation, uncertain reasoning and so on. The cloud model is the foundation of the cloud theory.This article improves the original interval cloud model based on the cloud model and the original cloud model by reducing the number of the numerical characteristics that expresses qualitative concept and defines new concept of the interval cloud model, and then builds up interval cloud generator. The interval cloud generator includes forward interval cloud generator, backward interval cloud generator, X-condition interval cloud generator and Y-condition interval cloud generator. At last, it is empirical analysis for the time series data by the interval cloud and the interval cloud generator.The data in this paper is the daily air temperature data of Chengdu from 2011 to 2014, and the data includes daily maximum air temperature and minimum air temperature. The data is from the website called ‘the weather report later’ and is absolutely true. The procedure of the empirical analysis for the time series data by the interval cloud and the interval cloud generator includes data pretreatment, creating the cloud model, activating predicted language rule and realizing forecast in the end. In the process of the empirical analysis for the time series data, the main tools include the cloud generator of the one dimensional cloud model and the interval cloud model and the language rule in uncertain reasoning theory. From the comparison between the predicted data and actual data, the empirical analysis for the time series data has good effects by the new interval cloud model. In this paper, all cloud generators and the graphs of the cloud model are made by statistical software called SAS.