A new compensation method for insulated core transformer power supply and its optimization using genetic algorithm

摘要

The insulated core transformer (ICT) power supply is employed as a stable and efficient high-voltage generator for the irradiation electron accelerator. However, insulation gaps in ICT would lead to the unequal output voltages of the rectifier disks, which might reduce the total output of the power supply with the same rectifier diodes and insulation distance. The paper presents a new full-parameter compensation method that adjusts the turns of the secondary windings and the values of compensation capacitors to decrease the nonuniformity of the output voltages of the each stages. Based on experimentally tested mutual inductance matrix, high-precision simulation models for the no-load and full load conditions are carried out. Furthermore, the genetic algorithm (GA) is chosen for the first time to deal with multi-objective optimization of ICT power supply system with the consideration of nonuniformity, load regulation, and winding current. The initial population is generated according to the existing compensation methods to improve the algorithm convergence rate. Finally, the best solutions show that the voltage nonuniformity is 0.57% and load regulation is 15.56%, which are much better than that of manual combined compensation method and capacitor compensation method. The results also confirm the high efficiency and advancement of the method in this paper.