Research on Fault Diagnosis Based on BP Neural Network Optimized by Chaos Ant Colony Algorithm
In view of shortcomings of BP neural network, which is slow to converge and tends to trap in local optimum when applied in fault diagnosis, an approach for fault diagnosis based on BP neural network optimized by chaos ant colony algorithm is proposed. Mathematical model of chaos ant colony algorithm is created. Real-coded method is adopted and the weights and thresholds of BP neural network are taken as ant space position searched by chaos ant colony algorithm to train BP neural network. Training result of chaos ant colony algorithm is compared with that of conventional BP algorithm and from both results it is can be seen that chaos ant colony algorithm can overcome the shortcomings of BP algorithm. It is proved that mathematical model of chaos ant colony algorithm is correct and optimization method is valid through experimental simulation for machinery fault diagnosis of mine ventilator.
Keywordschaos ant colony algorithm fault diagnosis BP neural network optimization algorithm
Unable to display preview. Download preview PDF.
- 4.Bio, S., Couzin, I.D., Del, N.B., et al.: Coupled oscillators and activity waves in ant colonies. Proceedings of the Royal Society of London Series b-Biological Sciences 266, 371–378 (1999)Google Scholar
- 8.Suddy, J.H.: An introduction to the behavior of ants. Arnold, London (1967)Google Scholar
- 10.Li, L.-X., Peng, H.-P., Yang, Y.-X., Wang, X.-D.: Parameter estimation for Lorenz chaotic systems based on chaotic ant swarm algorithm. Acta Physica Sinica (1), 51–55 (2007)Google Scholar