Mathematical modeling is a versatile and important tool in chemical process research and development. Using mathematical notation and mathematical logic to provide an abstract and rational quantitative description of the flow, transfer, and reaction processes in chemical unit operations, it can enable the underlying characteristics of chemical processes that are not visible to the naked eye and not detectable by existing sensors to be presented to researchers in a clear and systematic manner. Focusing on the design and optimization of olefin polymerization processes, the team has developed multiple types of mathematical models and simulation methods covering macroscopic reactor scale, mesoscopic flow structure scale, and microscopic molecular chain scale, with the objectives of reaction process enhancement and high-end polyolefin material manufacturing, liquid-solid, gas-liquid-solid, etc. Further, artificial intelligence techniques and advanced signal statistics have been introduced into the mechanism modeling to achieve more accurate portrayal and prediction of the essential characteristics of chemical processes at multiple scales. The research results have been funded by the Key Fund of National Natural Science Foundation of China and Integrated Project, and some of the results have been applied to the installations of many large petrochemical enterprises, such as Tianjin Petrochemical, Hainan Refining and Chemical, Zhongan United, Shenhua Ningcoal, etc., which provide important theoretical support for industrial reactor scale-up, large proprietary equipment manufacturing, and high-end polyolefin new grade development.