No abstract

The thesis describes the development of a dynamic model of a microturbine system. The thesis is done in close cooperation with Turbec AB and the model is adjusted, tuned and verified against their microturbine T100. The microturbine unit consists of a compressor and a turbine connected on a single shaft to a high-speed generator. Moreover there is a combustion chamber, a recuperator and a gas/water heat exchanger. A control system regulates the speed, the temperature and the electric power. To control the frequency, voltage and current of the outgoing power, the microturbine uses power electronics. The model is to be used in the research and development department at Turbec AB. Possible applications are in areas as control strategies, dynamic performance verification, operator training and control software/hardware verification. Since the potential applications are so different, the emphasis throughout the thesis has been on a general model that can be used in as many different operating ranges as possible. This should be done while retaining a high degree of accuracy at the most common scenario, running on full load. The emphasis has also been on the functionality and accuracy of the complete model over more detailed modelling of each component. In the thesis, the thermodynamic theory of each component is described and how it is modelled in Modelica. The microturbine model in the thesis covers the gas turbine unit, the control system and the mechanical part of the generator. The electric part of the generator and the power electronics are not included in the model, due to the limited time of the project. The thesis also discusses the well-known problems with modelling as e.g. discretisation and interpolation. The model uses components from the ThermoFluid library, the Modelica standard library and the Master's thesis of Perez (2001). Static verification has been done with data from DSA, a steady state calculation program, which accurately represents the real microturbine. The Modelica model developed in the thesis has been found very accurate at the main operation range (100 kW down to 50 kW), with an average error of 0.6 % of the 13 most important thermodynamic variables at full load, 100 kW. Dynamic verification in three different scenarios has been done against the real microturbine and the model shows a good fit to the measured data.

No abstract

Abstract-The implementation of anti-windup methods in mid-ranging control needs further attention. It is demonstrated how use of standard anti-windup schemes may give unnecessary performance degradation during saturation. The problem is illustrated for two separate systems, control of oxygen concentration in a bio-reactor and temperature control of a cooling system. In the paper, guidelines are derived for how to design the standard anti-windup scheme to recover performance. As an alternative a modified anti-windup scheme for midranging control is presented that minimizes the performance degradation during saturation.

No abstract