Multi-Domain Modelling - An Advantage In Modelling, Simulation And Control Education

The paper deals with an important aspect of continuous systems modelling and simulation approaches, with possibilities for multi-domain modelling. The traditional approach is based on block oriented schemes in which causal relations play an important role. However this causality is artificially generated in order to fulfil appropriate conditions for simulation on conventional sequential computers. Fortunately new concepts which are based on object oriented approaches, physically oriented connections and algebraic manipulations enable the so called acausal modelling which can efficiently be used for multi-domain modelling. The advantages and disadvantages of traditional and more advanced approaches are discussed. The described multi-domain approach is illustrated with modelling of a laboratory helicopter using a very popular multi-domain modelling environment Dymola with Modelica language. A multivariable controller was designed for a wide operating range. The control signals which drive the main and the tail rotors enable the desired pitch and rotation (azimuth) angles. The multidomain object oriented approach enables an efficient coupling of elements from mechanical libraries and a block library for the implementation of control system.