Study on Fault Diagnosis Based on the Qualitative / Quantitative Model of SDG and Genetic Algorithm

In term of multivariate operating conditions, complex dynamic performance and steady qualitative logic relation between variables in power plant thermal process, signed directed graph (SDG) is introduced to apply in fault diagnosis of power plant thermal system. SDG is a self-contained method to effectively diagnosis system failures, which can be constructed effectively by using the grading modeling method aided by simulation technology, but intrinsic limitations restrict it applies in fault diagnosis. Considering the relation among node of SDG can be effective described by constructing a qualitative and quantitative model; PCA that can monitor the correlation among different variables in the system and overcome the shortcoming of single variable analysis in determining the faulty node possibility, the genetic algorithm can be used to search possible fault propagation path quickly, a intelligent fault diagnosis approach is studied in thermal system field. The case studies show the qualitative and quantitative model of SDG has better resolution in fault diagnosis of power plant