Code No: 52214/MT
M.Tech. II-Semester Examinations, July/August-2006.
NEURAL AND FUZZY SYSTEMS
(Common to Power System Control and Automation, Power
Electronics, Electrical Power Engineering and Electrical Power
Time: 3hours Max.Marks:60
Answer any FIVE questions
All questions carry equal marks
1.a) Explain the structure of Biological Neurons.
b) Trace the historical development of Artificial Neurons.
2.a) Explain the significance of non-linear transfer functions.
b) Describe reinforcement learning with emphasis on Critical
3.a) Compare Perceptron learning rule and Delta rule.
b) Explain convergence of a network.
4.a) Discuss BAM with reference to simulated annealing.
b) How can Hopfield network be used in A/D converter?
5.a) Discuss the drawbacks of competitive learning.
b) Perform three iterations of Kohonen’s SOM to classify natural
numbers from 1 to 10 in to two groups as odd/even numbers.
Take neighborhood as 2.
6.a) If a bottle of water is labeled as “Probability of containing Potable
water is 0.5” and the other bottle is labeled as “Possibility of
containing Potable water is 0.5”, which is more preferable? Give
b) What are fuzzy cardinalities?
7.a) Explain Center of Area defuzzification method.
b) What id composition of rules?
8.a) Explain how LVQ can be used in fault diagnosis.
b) Discuss the application of fuzzy logic in control of a steam boiler.