A Neural Network Based Method for Cost Estimation 63/20kV and 132/20kV Transformers

Power transformers are one of the most important components in electrical network which has a vital role in electrification. Large numbers of transformers are required while expanding the electrification. In initial steps of evaluating project, having an estimated cost in short time can be beneficial. Around 15% of investment in transmission system goes towards transformer and since the major amount of transformers costs is related to its raw materials, cost estimating process can be an issue of crucial importance. In this paper, a new method is presented to estimate transformers pricing. In order to this aim, a unique Multi-Layer Perceptron (MLP) neural network has been designed for two various types of 63/20kV and 132/20kV transformers in Iran. In the next stages the cost of the transformer is estimated, finding suitable coefficients for the weight of copper, iron, transformer oil (that are MLP neural network outputs) and a constant coefficient that is related to manpower cost and other components of transformer costs. The requirement training data for MPNN are the obtained from the transformers made by Iran-Transfo Company during last 4 years.