Abstract: The process control system of the municipal solid waste incineration (MSWI) has characteristics of closed running and strict safety requirements. The key problems that need to be solved for the actual application of intelligent modeling algorithm includes how to realize the synchronization production of the offline multi-modal data and how to build a verification environment of data-driven prediction model like the actual industry. In this paper, a multi-modal data-driven verification platform for MSWI process is proposed, which consists of multimodal historical data synchronization subsystem and multimodal historical data-driven modeling subsystem. First, the structure of the simulating real-time verification platform is designed based on the abstract description of the predicting process of some key process parameters by domain experts. Then, multimodal data-driven prediction models for the furnace temperature, flue gas oxygen content and boiler steam flow are established. Finally, the hardware environment is built and the corresponding software system is developed. Thus, the cooperative operation between the platform subsystems is realized. Based on the actual process data and flame video, it is verified that the platform can solve the problems of difficult sampling, synchronization and matching in the construction of multi-modal data driven prediction models. Moreover, this study provides a reliable engineering verification environment.