Parameter estimation triggered by continuous-time parity equations

Faults which appear in technical processes can often be described as additive or multiplicative faults with respect to the residuals. To perform fast detection of these faults a continuous-time parity space approach combined with adaptive thresholds is used. Parameter deviations estimated directly from the residuals ease the identification of the fault and provide information about its size. The combined fault detection scheme is applied to a real permanently excited d.c. motor.