Feed-forward Control and Dynamic Modelling in Temperature Control of Buildings
This thesis is mainly about the investigations of different control strategies (and in particular feed-forward) in order to improve indoor climate and/or save energy in buildings. One part of the thesis deals with different ways to obtain dynamic models for climate systems in buildings. The main purpose of this part is to obtain tTansfer functions and state space models for buildings and using them for simulations in order to investigate different control strategies. In the major part of the thesis, different control strategies, with and without feed-forward, are investigated and the benefits of using feedforward in building control systems are discussed. Finally, in the last part of the thesis, the benefits and drawbacks of different types of controller are investigated and discussed. In this part, different methods for tuning of parameters of PIL) controllers are also compared.
The main conclusion, drawn from the simulations, is that a more extensive use of feedforward from intemal disturbances is veiy advantageous in many temperature-control applications. It gives better controller performance, and, at the same time, it will often reduce the energy use. From the resuits, it can also be conciuded that a change from an ON-OFF controller to a P-, PI-, or PIL) controller implies improvements both in energy efficiency and in controller performance. Another conclusion is that a change from the traditional Ziegler-Nichols tuning method to newer tuning methods improves both energy efficiency and control properties,
state space models
transfer function models