Modelling and control of energy and mass transfer in imperfectly mixed fluids

Some processes in several industrial and environmental systems, are the processes related to so called imperfectly mixed fluids: non-homogeneous gas and liquid mixtures with spatial gradients in energy variables, mass variables or impulse variables. In this work the accuracy of prediction of a model concept in modelling the dynamic behaviour of the micro-environmental factors temperature and humidity within an imperfectly mixed space is evaluated. In contrary to existing models in ventilation system studies this model concept: - is a model based on physical laws combined with a mathematical identification procedure, - is based on the assumption that a ventilated space is a completely non-homogeneous mixed air space, - assumes the existence of a time varying three dimensional air flow pattern, - gives the dynamic response of inside temperature and humidity to non-linear variations of the process inputs: ventilation rate and heat supply, - is a compact model that can be used and implemented for the application of model based control, - does provide a deeper understanding of the physical processes behind imperfectly mixed fluids because every model parameter has a physical meaning. To investigate the practical application and implementation of the model concept, the model is applied to two test installations: a ventilated space and small test container filled with water. In a first part of this work, a method is proposed to identify the unknown parameters as defined in the model concept. The developed identification method is then applied for estimation of these parameters using measurements as gathered on both test installations. After this model building process, the identified model is validated extensively. By comparing measured data with simulated results of inside temperature and humidity, it is shown that the model based on physical laws, combined with a mathematical identification procedure (a so called "grey box model"), is capable to model the energy and mass transfer to systems with different system dynamics, different flow conditions and on a different scale. A second part is devoted to the physical interpretation of the model parameters. In a last part of this work, both modelling approaches, respectively the model based on physical laws combined with a mathematical identification procedure and the black box model, are used as a basis for a model based predictive control system. It is shown by experimental results that the controller is capable to control the inside temperature, the fluid flow pattern and the distribution of temperature within the imperfectly mixed fluid.