Desiccant Air Unit (DAU) offers a complete control of air temperature and humidity in the conditioned space. Its key component is the desiccant wheel which provides the functions of air desiccation and regeneration. The aim of this study is to develop a methodology for obtaining a dynamic model of the desiccant wheel which can be used for the model-based control algorithms of DAU. The desiccant wheel can be regarded as a multi-input/multi-output (MIMO) system. The first part of the thesis is devoted to the modeling of the desiccant wheel based on energy and mass balance equations. The resulting set of equations is formulated as a second order state-space system without delay. The second part of this thesis concerns the experimental identification of the parameters of the state-space model of the desiccant wheel by using a black-box and a gray-box approach. In the case of the black-box, all the parameters of the model are identified experimentally. The identified parameters have values which minimize the difference between the output of the model and the experimental values. The parameters of the black-box model do not have physical significance. Although precise in the range of variation of the inputs in which the parameters were identified, this model gives significant errors in other domains of variation of the inputs. The parameters of the gray-box model are physically significant. Compared with the black-box models, the gray-box model was less accurate for the domains for which the parameters were identified, but it was notably more robust when applied to other ranges of the inputs. Since the parameters are related to physical properties, their values do not vary significantly with changes of the operating point used for identification. For the gray-box approach, the parameter values obtained for the linear models are almost identical for all local models on the desiccation side and all the local models on the regeneration side, suggesting that a local model may be valid for all the complete range of input variables. Using the above results, a final model of the desiccant wheel was developed, comprising two global models: one for the desiccation side and another for the regeneration side. The third part of the thesis deals with the identification of mass and heat transfer coefficients of the air within the desiccant wheel using a gray-box model. The mass transfer coefficient, the convective heat transfer coefficient and the Nusselt number were obtained by defining the variable parameters of the model as a function of a single variable and by expressing the constant parameters as a function of the geometric and material properties of the wheel. This work contributes to the development of a state-space model used for the synthesis of control algorithms for the desiccant wheel.
