Summary: | 碩士 === 國立臺灣科技大學 === 電機工程系 === 98 === At present, the design of transformer mostly used steady-state thermal model instead of transient thermal model to estimate temperature rise. For the particular application of isolated transformer in uninterruptible power system (UPS), the transient thermal constraint at specific time points needs to be regulated to meet the actual requirement. In view of this, this thesis is mainly concerned with the optimal design of UPS isolated transformer considering transient thermal constraint. First, a numerical transformer model taking transient thermal constraints into consideration is presented. The error between the actual measurement data with simulation results is analyzed and discussed to improve the estimation accuracy of the numerical transformer model. Then, incorporating with the aforementioned numerical model, the genetic algorithm (GA) is applied to the optimal design of UPS isolation transformer with minimum cost due to its excellent capability of obtaining the global optimal solution with higher probability and preventing form getting stuck in local optimum. The practical constraints to be considered include electrical characteristics, and the transient and steady-state temperature limits. Finally, taking an 80kVA UPS isolated transformer as an example, an optimal design with different efficiency restrictions has been conducted. Simulation results form the study show that the original design form the manufacturing company has been improved and the minimum cost can be achieved.
|