Abstract: Smart agriculture is an important direction for agricultural development. As a digital tool for accurate management and intelligent decision-making of crop production, crop growth model is one of the core technologies of smart agriculture, which is called smart agricultural brain. Here, I introduced the development history of crop growth models in recent decades, which included germination stage, initial stage, rapid research and development stage, deep development stage, improvement and application stage. The characteristics and limitations of several typical models (Wageningen series models, DSSAT model, APSIM model, STICS model, etc.) widely used in the world were emphatically introduced. There are many shortcomings in the application of current crop growth models, mainly manifested in the weak generalization ability and poor migration ability of crop growth models, which limited the regional application ability of the models. The mechanism of response of crop growth and development to environmental factors was not well understood. Meanwhile, the quantitative expression model needed to be improved. The lack of quantitative description of adverse effects such as extreme weather events, pests and diseases affected the simulation accuracy of the model. It was difficult to balance the complexity of the model and the convenience of application. The application of crop growth models was not sufficiently integrated with current new technologies. The insufficient interpretability of genetic parameters in crop growth models impacted the prediction ability of the models.

Key words: crop growth model, model development, model application, problem and discussion