Improving Core Quality in Power Distribution Transformers Using Machine Learning Methods

Abstract

The estimation of individual core losses of wound core power distribution transformers are particularly important since their core costs account for around 30% of their overall material cost and are one of the key determinants of their quality. In addition, accurate calculations of individual core actual losses are extremely difficult, since actual losses show a divergence of up to 20%, in relation to the theoretical individual core losses. This paper demonstrates the use of Machine Learning (ML) techniques, namely Decision Trees (DTs) and the Learning Vector Quantization (LVQ) neural network to the enhancement of each core's quality in wound core power distribution transformers. The DTs method makes use of inductive inference to automatically build decision rules and apply them to the power distribution transformers production procedure. In the LVQ neural network, any set of input vectors can be classified by using supervised training of competitive layers. Real industrial measurements were used to create the learning and test set. Information includes measurements of the production line's quality control as well as the electrical properties of grain-oriented steel. The resulting DTs present a success rate of 94%. Based on these DTs, rules comprising the most significant parameters and their threshold values can be derived. These are used to lower the actual losses of individual cores, hence raising their quality. The LVQ neural network approach achieves a total classification success rate of 95%.

Keywords

• Power distribution transformer
• Core quality improvement
• Losses
• Learning vector quantization
• Decision trees