Abstract: With the promotion of "retreat from city and move into park" policy and the rapid development of steam heating networks for industrial parks，the complexity of heating network dispatch has exceeded operators' response speed and judgment capability based on their experience.Thus，an optimized dispatch technical framework of an intelligent heating network in an industrial park is proposed.Then，a digital twin topology network model mapping to the actual heating network is established，and a simulation calculation method for the steam heating network is advanced as the prediction and analysis basis of the optimized dispatch. Taking the large-scale steam heating network of a printing and dyeing industrial park in Zhejiang Province as research object，the dispatch results after shutting down the heat sources made by empirical method were compared to that made by intelligent dispatch. The results show that the intelligent heating technology realizes the optimized dispatch of the steam heating network accurately. After the optimization，the maximum deviation between the expected temperature and the actual temperature of the heating network is 2.27%，while the maximum deviation between the expected and the actual pressure is 8.57%.The dispatch made by empirical method results in a 0.14 MPa drop of steam pressure at user end on average，while the average steam pressure at user end only reduces by 0.07 MPa by taking optimized dispatch.The results indicate that the optimized dispatch effectively reduce the adverse effect of dispatch scheme on steam consumption and intelligent heating technology is promising in industrial parks.

Key words: steam heating network, optimized dispatch, cyber-physical integration, intelligent heating, digital twin