Mathematical Model of Petroleum Consumption in Thailand

Based on historical data, the paper develops a mathematical model to describe characteristics of petroleum consumption in Thailand. The data is divided into 3 components: trend, seasonal and residue. The trend component is formulated as a transfer function model with various key economic factors as input of the system. The optimization technique based on sequential quadratic programming (SQP) is applied to identify the parameters of the model. The detrended data is characterized by its frequency spectrum then the development of the seasonal component follows. Finally, the Box-Jenkin approach is applied to the residue to develop a stochastic ARMA model. Key words: Petroleum consumption model, time-series analysis and component model