Research On Generalized Predictive Control Of District Heating Return Water Temperature

District heating system is a complex industry system, whether it’s running in science or not will impact the people’s livelihood, energy conservation and emissions reduction. Heat exchange station and building are as the heating terminal for the heat users, their running state has the direct influence on the heating quality. In currently, there are main problems such as:monitoring the heating parameters isn’t comprehensive, so it’s difficult to analyze the heating working state in detail; heating process has the great inertia and great time-delay features and so on, so it’s difficult for the normal PID and other algorithms to obtain a satisfied control effect.In this paper, the district heating system of an university in Dalian was selected as the research object, and the district heating network platform based on Internet of things architecture was also designed in this paper. Moreover, aiming at the problems for these controlled objects, like the heat exchange station and building’s time-delay and nonlinear features, the CARIMA model for heat exchange station and building’s heat transferring was established in this paper, and the GPC theory was applied into return water temperature control.Firstly, the district heating monitoring and management platform based on Internet of things was designed in this paper, the purpose of monitoring the field heating data in dynamic was achieved, and the whole heating data was also collected, it’s helpful for the managers can coordinate the every heat exchange station and building’s dynamic running better, it’s also convenient for the managers to give the reasonable heating operation guidance.Secondly, the mathematical model describing the heat exchange station and building’s heat transferring was established by the step response curve method, in order to meet the computer control’s demand, these mathematical models were discretized, and the accurate CARIMA models were also obtained, this can create conditions for the GPC algorithm’s implementation, and this can give the theoretical basis for the heating engineering model calculation.Thirdly, the heat exchange station’s secondary network return water temperature control and building’s return water temperature control were researched in simulink, the results indicated that compared with the normal PID algorithm and basic GPC algorithm’s disadvantages, the implicit GPC algorithm can have the better control quality and dynamic performance for the system output, this implicit GPC algorithm is suitable for the district heating system which is a great inertia, great time-delay, and uncertain dynamic feature system.Lastly, the implicit GPC algorithm can be applied into the hot water boiler supply water temperature control, so as to make this algorithm can be appropriate for the whole district heating system. It can provide the theory guidance for the heating optimization running, it can also provide the new approach and possible for improving the district heating quality.