The Applications of Neural Networks in the Engineering Cost of Transmission Line

Abstract

With the rapid development of the electric power industry in recent years, the strengthening of the power construction market and the diversification of the main body of power investment, there appears a prominent question in front of the project owners——How to control and reduce construction costs? There are many methods to estimate the cost quickly and accurately. Among the common methods and some new ways which have appeared in recent years, people can find about seven types out of them, in which, neural network model is known for its versatility and adaptability. It does not exclude new sample. On the contrary, it improves its ability to generalize and forecast with the increasing number of samples. Therefore this paper establish a cost estimation model by introducing neural network which is based on the optimization of genetic algorithm, and expresses the relationship implied in the interior of data by using the network topology and parameters by studying a large number of samples so as to fit the conventional non-linear mapping relationship between the amount and cost of a transmission line project. The results show that the artificial neural network model has a significant effect on the project cost estimation. The introduction of neural network model will certainly promote the development of informatization of power project costs management.

BP neural network construction cost prediction model based on different optimization algorithms

“…The principle of a Backpropagation (BP) neural network, also known as a multilayer perceptron (MLP), is to train a feedforward neural network using the backpropagation algorithm [25]. The BP neural network consists of multiple layers of interconnected neurons, including an input layer, one or more hidden layers, and an output layer [26].…”

Estimation of construction project building cost by back-propagation neural network

“…The experimental results showed that the method had a stable performance and could accurately estimate cost. Sun et al (2013) optimized the neural network using genetic algorithm and established a cost estimation model to realize the estimation of power engineering cost. Hsiao et al (2012) designed a new model to estimate the cost of semiconductor connection construction project by combining component ratio method, fuzzy adaptive learning control network (FALCON), fast chaotic genetic algorithm (fmGA) and three-point cost estimation method and verified through the study of 54 cases that the model had an estimation accuracy of 83.82 per cent.…”