Fault diagnosis based on relative transformation of information increment matrix
Chenglin Wen · Journal of Central South University(Science and Technology) · 2013
The existing information incremental matrix fault diagnosis methods can effectively overcome the shortcomings that traditional principal component analysis method cannot identify fault because of its serious pattern composition effects,but it ignores the impact of differences in scale for system variables,which often induces the important variables with small absolute value undetected when faults with smaller absolute values happened.While the small fault usually plays a key role for the security and stability of the system.Thus,according to the different important levels of different variables in the actual production process,on the basis of the relative transformation,a modified information incremental matrix fault diagnosis method was proposed.The simulation results show the effectiveness of the proposed method.The results show that this new method not only can keep the same detectability as the original system,but also can detect the fault of some small but important variables.