Design Optimization Performance Analysis And Cost Estimation of A Transformer Using Artificial Intelligence Techniques

Abstract

Materials and Methods: In this thesis, for constrained single-objective optimization, in addition to the conventional method, Genetic Algorithms involving three different selection operators, viz. and have been employed for minimizing active part cost of a transformer. newlineFor minimization of transformer losses and cost simultaneously, multi- newlineobjective optimization of transformer using and has been employed. Elitist non-dominated sorting and crowding distance are used to obtain pareto-optimal solutions. technique is then suggested for obtaining best compromised solution among non-dominated solutions. newlineFurther, annual load for three different locations in Bhuj city has been obtained from GETCO and the transformer design having the minimum has been suggested. newlineResults and Discussion: After applying and techniques for solving problem, it has been observed that and are able to find a better value of the objective function as compared to From statistical point of view, is found to be more robust as compared to and After application of for multi-objective optimization, technique enabled Decision Maker to select any solution posteriori, from the available non-dominated solutions. newline.Conclusion: After comparing the performance of and it has been observed that obtained cost saving of 2.73% and 1.95% for 1-star and 2-star rated transformers respectively, as compared to conventional method. newline For multi-objective optimization, has been able to obtain good diversity among pareto-optimal solutions and it can be inferred that it is possible to reduce active part cost and load losses simultaneously, at the expense of slight increase in no-load losses. newline