Non-conventional Transformers Cost Estimation Using Neural Network

Since the cost of transformer can be divided into 50-60% for material, and the rest being labor costs and modest profit, therefore as the major amount of transformers costs is related to its raw materials, so it has a high importance in costs estimating process. This paper presents a new method to estimate transformers pricing. The method is based on multilayer perceptron neural network (MPNN) with sigmoid transfer function. The back-propagation (BP) algorithm is used to adjust the parameters of MPNN. The required training data for MPNN are the obtained information from the transformers made by Iran-Transfo Company during last 4 years. A Multi-Layer Perceptron (MLP) neural network has been designed for 132/33KV transformers (which is classed as non-conventional in Iran). By finding suitable coefficients for weight of the copper, iron and transformer oil (that are MLP neural network outputs) and a constant coefficient that is related to manpower cost and other transformer components costs, the cost of transformer is estimated.