Multiple ANNs combined scheme for fault diagnosis of power transformers

It is very important to find the incipient faults of power transformers for keeping away from the further deterioration in modern power system. For enhancing the accuracy of fault diagnosis for power transformers, a multiple ANNs scheme is proposed in this paper. In this scheme, a novel ANN method ELM is used to establish the base classifier for its good performance and fast learning speed. Secondly, the several base classifiers based on single ELM will be combined by consulting ensemble techniques. And then a multiple ELMs method is obtained. The real gas records data from a power company is used to establish fault diagnosis system for power transformers based on the new multiple ELMs method. For comparison, the conventional method and other ANN method are used to build fault diagnosis models by the same data. The experiments demonstrate the new multiple ELMs method has the best performance in both learning ability aspect and generalization ability aspect for fault diagnosis of power transformers.

[1]  Yoav Freund,et al.  Experiments with a New Boosting Algorithm , 1996, ICML.

[2]  Guang-Bin Huang,et al.  Extreme learning machine: a new learning scheme of feedforward neural networks , 2004, 2004 IEEE International Joint Conference on Neural Networks (IEEE Cat. No.04CH37541).

[3]  Chee Kheong Siew,et al.  Extreme learning machine: Theory and applications , 2006, Neurocomputing.

[4]  Yoav Freund,et al.  A decision-theoretic generalization of on-line learning and an application to boosting , 1997, EuroCOLT.

[5]  Durga L. Shrestha,et al.  Experiments with AdaBoost.RT, an Improved Boosting Scheme for Regression , 2006, Neural Computation.

[6]  Yoav Freund,et al.  A decision-theoretic generalization of on-line learning and an application to boosting , 1995, EuroCOLT.

[7]  Zhi-Zhong Mao,et al.  An Ensemble ELM Based on Modified AdaBoost.RT Algorithm for Predicting the Temperature of Molten Steel in Ladle Furnace , 2010, IEEE Transactions on Automation Science and Engineering.

[8]  Geoffrey I. Webb,et al.  MultiBoosting: A Technique for Combining Boosting and Wagging , 2000, Machine Learning.

[9]  Yann-Chang Huang,et al.  Evolving neural nets for fault diagnosis of power transformers , 2003 .

[10]  Thomas G. Dietterich.  An Experimental Comparison of Three Methods for Constructing Ensembles of Decision Trees: Bagging, Boosting, and Randomization , 2000, Machine Learning.

[11]  Xiao-Yun Sun,et al.  Application of Combinatorial Probabilistic Neural Network in Fault Diagnosis of Power Transformer , 2007, 2007 International Conference on Machine Learning and Cybernetics.

[12]  Leo Breiman,et al.  Bagging Predictors , 1996, Machine Learning.