Fault Detection using Dynamic Parity Space Approach

This paper review the effectiveness of the parity space approach to identify faults or disturbance in a system. The most commonly used is the observer based procedures, and redundancy relationship method. This involves analytical mathematical analysis of geometry and bilinear algebra. Then, technological advances which require complex computation such as artificial intelligence and genetic algorithm had made tremendous improvement to fault Detection and Isolation (FDI) analysis. Dynamic Parity Space Approach was studied for a discrete state-space model. Important data will be extracted using this approach especially for residual generation which is the backbones of FDI analysis. Subsequently, at each time instant k, the generated residuals will form a matrix that will define the fault signature. It is remarkable that this approach is proven in this study to be effective in diagnosis and faults isolation.