Estimation of ARX parametric model in regional economic systems

Abstract

The purpose of this paper is to show how an economic system model can be selected that would allow a regional authority to control the economy of a region. This paper presents a new idea of system identification method based on fuzzy evaluation logic. It applies parametric estimation in analyzing the data of the inputs and the outputs in a nonlinear system, and then selects a regional economic system model from several nonlinear models by means of the system identification method. Due to the lower explanatory power of the present models, this paper focuses on the shortcomings of the present model based on a simple mechanism methodology which has a complicated structure and lots of variables. An alternative is to construct an ARX model with systematic consideration. The advantage of ARX model is simple, with few variables as well as being actually data-based. The result is that ARX117 model is the best one, and the IV Estimation Method is the one we suggest for use. Since this ARX model can reflect the complex characteristics of a regional economic system by integration of experts' knowledge, we can design the most reasonable fuzzy controller for a regional economic system. © 2007 Wiley Periodicals.