Intelligent method of reducing BIT´s false alarm based on SVM_FCA_HMM

BIT´s false alarm is one of the most important reasons of restricting testability´s development. In order to decrease BIT´s false alarm and develop BITE´s performance, this paper discusses present situation of BIT´s false alarm at first, then introduces several advanced models of reducing false alarm which are neural network, support vector machine, fuzzy clustering analysis and hidden Markov model. After then, this paper compares the advantages and disadvantages of these models about reducing BIT´s false alarm. Synthesizing the advantages of SVM, FCA and HMM, this paper gives an intelligent model based on SVM_FCA_HMM to decrease false alarm, and analyzes theoretically this model´s functionary and efficiency. Finally, this paper applies the method in the aero-engine´s digital electronic control system and discusses its rationality and applicability, and proves that this model can be applied generally.