Model-Based Fault Diagnosis with Fractional Models

Paper presents the results of experimental research of fault detection based on difference signal of plant and model controllers output. The plant was simulated by electronic circuit with supercapacitor 'damaged' by discharging it by the current source. The model of the plant was described by fractional order differential equations. Such approaches allowed for some processes and the newest technologies, obtain better modelling accuracy. Model parameters was derived used identification procedures based on step response signal. The fault detection problem of this paper can be used as simple benchmark example to test new fault detection rules before applying it to real systems.