Abstract: “The development of model-based fault detection methods for engineering systems requires a high degree of consistency between the behaviour of mathematical model and the actual system in order to produce reliable results. It means that the mathematical model must be very accurate. Due to the usually complicated equations describing the dynamic behaviour of these systems, some parameters can be roughly estimated based on the steady-state analysis. It is necessary to use alternative approaches that optimise the parameter values and make them more accurate. This paper presents a parameter validation approach that can be used in the case of incomplete information of the system state and improve the quality of a model. The approach is used for the determination of the acceptable deviation limits between model and system.”

Keywords: Modelling, Simulation, Fault detection, Parameter Validation.