Abstract In today’s increasingly globalized world, trade between countries has become more interconnected, and international trade has emerged as the primary driver of economic development for all nations. This paper proposes a standard trade model for the new situation using the Dixit-Stiglitz model. Starting from the consumer’s point of view, the two-stage maximum method is adopted. Using the Lagrange function, the demand utility function of consumer preferences is processed, and the model optimization scheme is deduced. The ARMA model is used to predict the volume of merchandise foreign trade, and the prediction range includes autoregressive and moving average processes. The indicators are used to evaluate the smoothness of the series within the sample period. A suitable sample of international trade data is selected. The results of the analysis are as follows: in the logistic autoregressive analysis test, the significance of all factors, except for the language and religion factors, passes the significance test of 0.001, and at the same time, the model goodness-of-fit is greater than 0.1, ranging from 0.115 to 0.175. In the analysis of the spatial evolution pattern of international economic trade and trade dynamic development, the spatial Gini coefficients of the first category of goods in 2000, 2014, and 2019 and the third category in 2000 are 0.4306, 0.4717, 0.4956, and 0.4673, respectively, which are lower than 0.5, showing a more balanced evolution pattern.