Coordination Fault Diagnosis Based on RBF

Abstract

In large-scale complex system, The establishment of a fast, accurate fault diagnosis system is more difficult because there exist many uncertain elements between the fault cause and the fault sign .A fault diagnosis system is established based on RBF cloud neural network ,the RBR (rule-based reasoning) and the CBR (case-based reasoning).The fault diagnosis system not only has the advantages of self-learning, high accuracy, randomness, fuzziness, etc ,and has the advantages of independently of mathematical model ,rich knowledge representation, mighty problem solving ability, etc. Theoretical analysis and simulation results show that the system is feasible and effective for fast and accurate fault positioning of complex systems.