Modeling of central heating system is an important basis for intelligent management of heating system. In this paper, eight heat exchanger stations in a branch of a pipe network were studied, PostgreSql was used to establish the information database of physical network and internet of things, and BP neural network was used to predict heat load. In the meanwhile, fluid simulation software Flowmaster was used to build simulation system, then coupled with the graphical user interface (GUI) in MATLAB to realize the dynamic visualization of operation conditions of the simulation system. The results show that the related coefficients R2 of heat load forecasts of the eight heat exchanger stations are all greater than 0.85 and the relative errors of temperature and pressure of supply and return water in simulation system are less than 10%; in addition, the dynamic visualization of operation conditions of the simulation system is achieved. The work and results of this paper preliminarily prove that the combination of the basic functions commonly used in Flowmaster and the graphical user interface of Matlab can economically achieve the operation simulation of heating system and the dynamic visualization of hydraulic and heat conditions. It is of great significance to generate new levels of the heat-supply network’s intelligence and realize the intelligent heating with functions of intelligent diagnosis, intelligent surveillance and multidimensional service push.
Keywords Flowmaster, PostgreSql data base, BP neural network, Matlab GUI, Dynamic visualization interface