Search Results

Fault detection and identification system generally uses state observers to generate redundant signals. State observers need sensor measure as input signal but sensor measure contains noise and this noise can damage redundant signals. In this paper, we will show how the sensor noise can affect redundant signals generated by a Luenberger observer designed to be used in a DOS structure. PI Observer due to its conception can be designed to be robust to sensor noise. We will propose a method using the recursive definition of the determinant of a matrix to design the PI Observer to be used in a DOS structure. Finally, we will show the performance of the PI observer for detecting and identifying faults in sensors with noise in its measurement.

State observers employed in the analytical redundancy approach can generate redundant signals. As the sensor has noise in its measures, this noise also affects the redundant signal. In this paper, we will show how the noise in the sensor signal can affect the redundant signal, generated for a bank of observers in a DOS structure, and some simulations results of this signal and finally we will do some considerations in the state observers design to reduce the noise level at the redundant signals.

Several papers presents fault detection and isolation techniques for fault in only one sensor; in this paper we will present a technique for multiples faults detection and isolation in sensors of dynamic systems. Multiples faults have less probability to occur but it is not null. So in critical applications the system needs to be operational even in this situation. In this paper we will present a design for a Multiples Faults Detection and Isolation (MFDI) system, an example to illustrate this technique and its respective results.

Eigenstructure techniques allow to detect and isolate faulty components in a dynamic process, such as sensor biases, actuator malfunctions, changes in dynamic parameters due to leaks and deterioration. Fault detection is the first step to achieve fault tolerance, but for this the redundancy has to be included in the system. This redundancy can be either by hardware or by software. In situations in which it is not possible to use hardware redundancy only the software redundancy can be used. Therefore using eigenstructure techniques, for the fault detection and isolation, the tests can be done through the angle between the residue vector direction and the fault direction vector. By this way, we can reduce false alarm and the alarm loss rates due to the noise and changes in system parameters.

This paper presents some techniques for fault diagnosis in aerospace and automotive systems. A diagnosis technique is an algorithm to detect and isolate fault components in a dynamic process, such as sensor biases, actuator malfunctions, leaks and equipment deterioration. Fault diagnosis is the first step to achieve fault tolerance, but the redundancy has to be included in the system. This redundancy can be either by hardware or software. In situations in which it is not possible to use hardware redundancy only the analytical redundancy approach can be used to design fault diagnosis systems. Methods based on analytical redundancy need no extra hardware, since they are based on mathematical models of the system.