Urban planning plays an important guiding role in land use change. The planning at all levels supports and complements each other, and influences the evolution of land-use pattern from top to bottom. Vector cellular automata, which takes irregular geographical entities as its basic units, can express the objective and complicated urban land use structure in a more realistic way, and has become one of the focus in the land use change research. However, when we face urban planning with the characteristics of hierarchical synergy, spatial guidance and controlled conductivity, some key issues are highlighted, such as the establishment of CA hierarchical system, the construction of hierarchical synergy and the acquisition of transition rules. After constructing the multilevel vector CA model, we chose Jiangyin City as the study area and its land use data covering the years 2007, 2012 and 2017 as the basic data to simulate the land use changes from 2012 to 2017. On the basis of multilevel vector cellular automata model, through the comparative analysis between the simulation results and the current situation of land use in 2017, the individual parameters of the model were modified to further improve the feasibility and applicability of the model, and then the urban land use pattern in 2022 was predicted. The simulation results indicated that the development of construction land in Central District has become saturated, while the expansions of construction land in Chengnan District, Chengdongnan District and Chengdong District were obvious. From an overall perspective, there was a tendency to gradually form a three-level urban-rural spatial settlement system "central city area-urban agglomeration-village". Using FoM index, we concluded that the simulation results had great simulation precision and the model had high feasibility and applicability, because the FoM values of the whole city and each district were almost greater than or close to 0.21. This shows that the simulation results are more accurate and the model is more effective in the simulation of land use change based on urban planning.