Study on Contamination Viscosity in Batch Transportation Based on the BP Neural Network

In view of deficiencies in current contamination viscosity calculation, the forecasting model of contamination viscosity is set up respectively on three different contaminations based on the analysis of the basic principle of forward back propagation (HP) neural network. The three BP neural networks are trained and simulated respectively using practical measure data. The results show that error ratios of three different contaminations are less than 2. 5%. It also indicates that the present method has higher accuracy and wider applicability than that of KerndaLMunnloe formula and Zdanowski formula proposed by Soviel scholars and completely meets the requirements of practical engineering.