Preparation Of Heat-resistant Coated Iron Yellow Pigment And Prediction Of Heat-resistant Color Difference By Neural Network

The heat resistance of iron oxide yellow pigment is poor and its application in plastic processing and coil coating is limited.In this paper,ferric oxide yellow was used as the precursor material,the surface of ferric oxide yellow was coated by means of mixing,monolayer and multilayer coating to improve its heat resistance.By changing the core factors such as reaction time and temperature,the optimal conditions of coating reaction were determined.Conventional pigment preparation and chromatic aberration test by chemical experimental means are time-consuming and laborious,and the accuracy is low for developing heat-resistant iron yellow pigment products.The iron oxide yellow production enterprise urgently needs an artificial intelligence method to quickly and efficiently automatically obtain the heat resistance color difference value of the coated iron yellow,and the neural network technology(BP)can realize the rapid and efficient test of the heat resistance of iron oxide yellow pigment.The main research contents and innovations are as follows:(1)In this paper,iron oxide yellow was used as the precursor and coated with calcium phosphate liquid phase precipitation method to improve its heat resistance.By changing the reaction time,reaction temperature and other factors,the suitable coating reaction conditions were determined.The microstructure of the pigment was characterized by powder X-ray diffraction(XRD),thermogravimetric differential thermal analysis(TG-DTA),infrared spectroscopy(FT-IR)and scanning electron microscopy(SEM).The results show that a layer of inorganic calcium phosphate film is coated on the surface of iron oxide yellow,and the coating layer is amorphous.The optimum reaction conditions were obtained as reaction temperature 50? and stirring time 5hours.After mixing calcium phosphate coated iron oxide yellow pigment with iron phosphate,the best mixing proportion of iron phosphate was deduced according to the color difference formula of heat resistance theory through the mechanism of heat replenishing iron phosphate,which can effectively improve the heat resistance of coated iron oxide yellow pigment.(2)Based on BP neural network,the rapid and accurate prediction of color difference of monolayer calcium phosphate coated iron oxide yellow pigment at 260? is realized.Combined with genetic algorithm,the deviation of BP neural network prediction is expected to be reduced and the accuracy of prediction is improved.The BP neural network(GA-BP network)optimized by genetic algorithm was applied to predict the thermal color difference of monolayer calcium phosphate coated iron oxide yellow pigment.The reliability of the neural network method was evaluated by comparing the difference between the predicted thermal color difference and the experimental value at 260?.Finally,the optimized model was used to predict the multi-coated iron oxide yellow pigment.The prediction results show that the ga-BP neural network color difference prediction model has better performance,and the relative error of prediction is about 10%,which basically meets the needs of the actual coated iron yellow pigment product development and production.In conclusion,GA-BP neural network can predict the color of coated iron yellow,and realize the rapid verification of experimental process parameters.It provides a new scheme for the research and production of coated iron oxide yellow pigment and lays a foundation for the accurate prediction of color difference of iron oxide pigment by neural network.