Engineering-oriented Modeling Of Power Transformers Under Dc-biased Magnetization

The dc-biased magnetization is an unusual exciting condition for the AC power transformers. The use of an iron core under dc-biased magnetization generates a dc component in the exciting current. The magnetic flux, the peak value of the exciting current increases significantly and the waveform of the current is seriously distorted with a lot of harmonics. The iron core is asymmetrical saturated and the leakage flux of the iron core is increased. The heavy noise, the serious vibration, and the local loss concentration can all occurre in dc-biased transformers. Moreover, the harmonics will cause the misoperation of the protective relay which is harmful for the safe operation of the transformers and the power system.The experimental research and simulation analysis of power transformers under different dc-biased magnetic field strength were completed according to the Natural Science Foundation of Hebei Province under Grant E2006001036. The measured induced voltage and exciting current was processed to obtain the electromagnetic properties of the iron core. A neural network hysteresis model was proposed to predict the hysteresis loops under dc-biased magnetization. The 3-D finite element model was presented to analyze the distribution of the magnetic flux and iron loss inside the iron core of dc-biased transformers. The main points of this paper are:1. A practical experimental approach was proposed to invesgate the eletromagnetic characteristics of dc-biased transformers. The technical requirements for the dc bias experimental resreach were determined. And the measurements based on the laminated core model of product level were carried out.2. The magnetic properties, such as the B-H curve,Ф-I curve and the loss curve, of the laminated ferromagnetic materials were measured and compared with those obtained under"standard"condition. This is essential in the 3-D finite element simulation of the transformers under dc-biased magnetization.3. Based on the obtained material property database from experiments, a neural network hysteresis model was established to predict the hysteresis loops under different dc-biased magnetic field strength. The approach for calculation the exciting current of biased-transformer was presented, in which the predict hysteresis loop corresponding to the specific working condition is used.4. An engineering-oriented method was presented and developed based on the MATLAB commercial software to compute the magnetizing current and the dc magnetic flux in the iron core. The comparison between the calculated results and the measured ones was carried out to validate the effectiveness of the proposed method.5. Two experimental schemes for how the dc-biased magnetic field strength applied to the transformer were proposed. The differences and relations between the two schemes were examined, that is useful for the establishment of the magnetic analysis model.6. A practical method was proposed to simulate the distribution of magnetic flux and iron loss inside the laminated core under dc-biased excitation, which is based on the 3-D electromagnetic field analysis software, MagNet. The good agreement between the calculated results and the measured ones proves the effectiveness of the method.