Nonlinear Support Vector Machines for Solving the PMC-Based System-Level Fault Diagnosis Problem

This paper deals with the system-level fault diagnosis problem which main objective is to identify faults, in particular permanent ones, in diagnosable systems under the PMC model. The PMC model assumes that each system's node is tested by a subset of the other nodes, and that at most t of these nodes are permanently faulty. Tests performed by faulty nodes are unreliable, and hence, they can incorrectly diagnose fault-free nodes as faulty or faulty ones as fault-free. In this paper, we describe a new nonlinear support vector machines-based (SVMs) diagnosis algorithm, which exploits the off-line learning phase of SVMs to speed up the diagnosis algorithm. The novel diagnosis approach has been implemented and evaluated using randomly generated diagnosable systems. Results from the thorough simulation study demonstrate the effectiveness of the nonlinear SVM-based fault diagnosis algorithm, in terms of diagnosis correctness, latency, and scalability. In addition, extreme faulty situations, where the number of faults is around the bound t, and large diagnosable systems have been also experimented to show the efficiency of the new nonlinear SVM-based diagnosis algorithm.