Utilized degradable data of coal-filled films from the accelerated UV chamber ageing degradation experiments, and on the basis of control factors’ analysis, presented a predicting model on degradable properties of th...Utilized degradable data of coal-filled films from the accelerated UV chamber ageing degradation experiments, and on the basis of control factors’ analysis, presented a predicting model on degradable properties of this film in photo-degradation according to back-propagation artificial neural network (BP ANN). 4 controlling factors in films degrada-tion, including temperature, the time of UV irradiation, the concentration and the type of coals were used as input parameters in the ANN model. While the degradable properties after film degradation, including the mechanical properties and carbonyl index, were used as output parameters. It was carried out by the neural network toolbox of Matlab 6.5 soft-ware and Visual Basic 6.0. Discussed partition of sample data and model’s parameters, and then selected the best configuration of ANN network. The accurate scope of predicting results was analyzed. This model has a high precision in predicting on properties of the coal-filled film degradation.展开更多
基金Supported by the National Natural Science Fund ( 20276056)Special Fund of Education Department of Shaanxi Province (03JK190)
文摘Utilized degradable data of coal-filled films from the accelerated UV chamber ageing degradation experiments, and on the basis of control factors’ analysis, presented a predicting model on degradable properties of this film in photo-degradation according to back-propagation artificial neural network (BP ANN). 4 controlling factors in films degrada-tion, including temperature, the time of UV irradiation, the concentration and the type of coals were used as input parameters in the ANN model. While the degradable properties after film degradation, including the mechanical properties and carbonyl index, were used as output parameters. It was carried out by the neural network toolbox of Matlab 6.5 soft-ware and Visual Basic 6.0. Discussed partition of sample data and model’s parameters, and then selected the best configuration of ANN network. The accurate scope of predicting results was analyzed. This model has a high precision in predicting on properties of the coal-filled film degradation.
