Pipeline transportation expense is predicted using the method of artificial neural network for the first time, Three-layer BP neural network model with 1-7-1 structure is set up. Input-layer's node is pipeline transportation flux, output-layer's node is pipeline transportation expense. The 25 study samples' training is completed by using the improved BP algorithm. Then the 6 samples are tested using trained network model, the error of predictive value range is within 4%. It completely satisfies engineering practical need. The model needs fewer parameters and can predict facultative transportation flux's pipeline transportation expense. Thereby it provides decision-making's gist for energy resource manage department making energy resource dissipative ration and planning finance department predictive cost.
Pipeline transportation expense predicting based artificial neural network
2003
4 Seiten, 4 Quellen
Conference paper
English
Minimising network management expense
British Library Online Contents | 2004
Optimization of the transportation expense of a firm with contractual supplies
Online Contents | 2000
|MANUFACTURING EXPENSE CONTROLS
SAE Technical Papers | 1960
|