Quantitative assessment of drought risk is of great significance for economic and social prevention and control management. In response to the problem of the wide peak distribution characteristics of the difference coefficient value and its dynamic characteristics as it changes with the evaluation sample value, based on the micro level migration of the connection components in the semi partial subtraction set pair potential, the trapezoidal fuzzy number dynamic value is used to characterize the uncertainty and continuous change process of the difference coefficient. The trapezoidal fuzzy number is simulated using random simulation method to quantitatively describe the evaluation results in the form of confidence intervals, a drought risk assessment model based on subtractive set pairwise potential trapezoidal fuzzy number random simulation is constructed, and an empirical study on dynamic drought risk assessment is conducted in Suzhou City, comparing and analyzing with existing methods. The results show that the negative development trend of drought risk in Suzhou City increased in intensity from 2007 to 2010, and decreased in intensity from 2011 to 2017. The risk level remained above level 2 for a long time and was in a state of partial drought. 2010 was the most dangerous year; the overall fluctuation range is relatively small and there is a trend towards positive development. It is identified that the risk subsystem and drought resistance subsystem are the main factors leading to the drought risk in Suzhou City. This method has strong explanatory principles, reasonable and accurate results, and the evaluation results are expressed in the form of confidence intervals, providing a new and effective approach for dynamic analysis and quantitative assessment of drought risk.