As the scale of the chemical park industry continues to expand, the impact of flooding on the park’s people and surrounding environment increases. This paper uses project scheduling theory to optimize the emergency rescue process in order to alleviate the suffering of affected people, promote the sustainable development of society and the environment, and take into account the characteristics of the dynamic evolution of flooding in chemical parks and the periodic renewal of emergency resources. We constructed a proactive–reactive multi-mode emergency rescue integrated scheduling model that aims to minimize the loss of affected people in the early stage of flooding and minimize the sum of the total deviation of the start time and end time of activities before and after reactive scheduling in the later stages of flooding. Furthermore, an ant colony algorithm was designed to solve the constructed model. Next, the effectiveness of the proposed model and solution algorithm was verified using simulations of actual cases. The calculation results show that using proactive–reactive integrated scheduling can improve the efficiency of emergency rescue and reduce the loss of affected people. Moreover, if a multi-mode rescue strategy is adopted, emergency rescue scheduling under four different resource combinations can reduce rescue duration and loss of affected people. The model can provide a decision reference for sustainable emergency rescue scheduling in chemical parks during a flood.