The use of complex and large systems and infrastructure is vital to carry out various activities in a country, and observance issues of passive defense in critical situations that could provide complete or partial services, is considered one of the evaluation criteria for these systems. Modeling and simulation of these systems to identify bottlenecks is important. Fault occurrance seems natural despite various provisions such as fault forecasting , fault prevention, fault coverage and fault tolerance. Artificial neural networks as a method of modeling and simulation, have many applications in monitoring complex and critical systems. Since artificial neural networks have been designed based on natural neural networks model, that possess the inherent capability of fault tolerance. Therefore, they must be able to take advantage of fault tolerance ability. This article presents a method to enhance the fault tolerance and improve neural networks based on Triple Modular Redundancy (TMR). It shows that, based on this technique, the desired fault tolerance has been favorably increased.
Hasani Ahangar, M. R., & Akhzami, M. (2013). Fault Tolerance in the MLP Neural Networks Using Triple Modular Redundancy. Passive Defense, 4(1), 51-61.
MLA
M. R. Hasani Ahangar; M. Akhzami. "Fault Tolerance in the MLP Neural Networks Using Triple Modular Redundancy". Passive Defense, 4, 1, 2013, 51-61.
HARVARD
Hasani Ahangar, M. R., Akhzami, M. (2013). 'Fault Tolerance in the MLP Neural Networks Using Triple Modular Redundancy', Passive Defense, 4(1), pp. 51-61.
VANCOUVER
Hasani Ahangar, M. R., Akhzami, M. Fault Tolerance in the MLP Neural Networks Using Triple Modular Redundancy. Passive Defense, 2013; 4(1): 51-61.
)