International Journal of Electrical and Computer Engineering (IJECE) Vol.1, No.2, December 2011, pp. 102~109 ISSN: 2088-8708  102 Received Sep 11 th , 2011; Revised Nov 11 th , 2011; Accepted Nov 17 th , 2011 Design Optimization of High-Frequency Power Transformer by Genetic Algorithm and Simulated Annealing 1 Amit Kr. Yadav, 2 O.P.Rahi, 3 Hasmat Malik, 4 Abdul Azeem Electrical Engineering Department, National Institute of Technology Hamirpur, H.P. India e-mail: amit1986.529@rediffmail.com Abstract This paper highlights the transformer design optimization using genetic algorithm (GA) and simulated annealing (SA). Any optimization problem, a given objective function is to be minimized keeping in view the constraints. Similarly, transformer design optimization problem involves minimizing the total mass (or cost) of the core and wire material by satisfying constraints imposed by international standards and transformer user specification. The constraints include appropriate limits on efficiency, voltage regulation, temperature rise, no-load current and winding fill factor. The design optimizations seek a constrained minimum mass (or cost) solution by optimally setting the transformer geometry parameters and require magnetic properties. This paper solves the said design problem by using genetic algorithm (GA) and simulated annealing (SA) techniques. The results of geometric programming (GP) technique have been compared with the results obtained by applying GA and SA techniques. It is quite evident from the results that the dimensions as well as mass of copper and core have been reduced in comparison to GP using same set of constraints. Therefore, the paper presents improved design of power transformer by these two techniques. The results of GA and SA have been obtained using optimization tool box MATLAB Release 9.1 which have not been applied for power transformer design so far. First it provides efficient and reliable solution for the design optimization problem with several variables. Second, it guaranteed that the obtained solution is global optimum. Hence paper demonstrates a better and efficient solution to high frequency power transformer design. Keywords: optimization, power transformer, genetic algorithm, simulated annealing technique, geometric programming 1. Introduction In today era of competition optimal design plays an important role. Engineering optimization helps in achieving optimal performance, cost competitiveness and optimal efficiency of any equipment/apparatus. In the problem of power transformer optimization mass of core and copper are the main materials for any transformer manufacturing have been optimized using GA and SA.Since dimensions of transformer core and winding are complex system and in conventional optimization technique such a problem formulation involves large number of variables making transformer design optimization complex. The objective of transformer design optimization (TDO) is to optimize mass of core and mass of copper subject to constraints imposed by international standards and transformer user specification. The transformer design is worked out using various methods based on accumulated experience. Transformer design methods vary among transformer manufacturers. While designing a transformer, much emphasis should be placed on lowering its cost by saving materials. The design should be satisfactory with respect to dielectric strength and mechanical endurance, and windings must withstand dynamic and thermal stresses in the event of short-circuit. In order to meet the above requirements, the transformer designer should be familiar with the prices of basic materials used in the transformer This paper presents a transformer design methodology based on an artificial intelligence technique. In case of a power transformer design, a design engineer has to consider several aspects of the transformer design such as core shape, size, properties, copper wire size etc and arrives at an optimum design. Improper design results in under utilization of materials. The study of designing power transformers using a computer was pioneered by [1] [2] and [3]. Later, [4] suggested a method for obtaining an optimized design of power transformers. However, procedure was not general and cannot be used conveniently for all power transformer design problems. Several other techniques were also used for design of power transformer [5]-[11]. In [12], attention was focused on a class of single-phase transformers that are employed when a high voltage withstand test is required for verifying the quality of the dielectric insulation of a given component. 2. Minimum mass Design of A High-Frequency Transformer 2.1 Nomenclature p E rated primary voltage (V, rms sine wave); s E rated secondary voltage (V, rms sine wave); S rated apparent power rating (VA); pf rated power factor a T ambient temperature ( C Ο ); m η maximum allowed efficiency; brought to you by CORE View metadata, citation and similar papers at core.ac.uk provided by Institute of Advanced Engineering and Science