Development On Simulating And Forecasting System Of Double Glow Plasma Surface Alloying Based On Neural Network

The Double glow plasma surface alloying has a complex process and a lot of technological parameters. Because of the non-linear relationship between the processing parameters and the surface alloying effect, it's hard to build a mathematic model for reflecting the rules of the technology, and thus it brings some problems to the research of double glow plasma surface alloying. Considering these characters, after having explored and analysed the technique, process and influencing factors in this field, this paper applied Artificial Neural Network and Expert System in analyzing, simulating and predicting the whole process. Finally, the Simulating and Forecasting System of Double Glow Plasma Surface Alloying has been built.The Simulating and Forecasting System based on Artificial Neural Network, Expert System and Database, comprises of client layer and GUI client interface, application layer developed by JAVA, and background database server using SQL Server. It includes three modules: Artificial Neural Network modeling and simulating platform, Expert System, technique ayalysing and predicting module. As a result, it implements some functions such as introducing the theory, analysing the parameters, simulating and predicting the technique etc. This paper applied the double glow plasma surface alloying Expert System to analyse the technique of double glow discharge plasma aluminizing on 45 Carbon Steel, then the mathematic model was built on Artificial Neural Network modeling and simulating platform. The simulation results and curves demonstrate that the Simulating and Forecasting System has a higher accuracy in ayalysing, simulating and predicting the process of double glow discharge plasma aluminizing on 45 Carbon Steel, and it is also indicated that this system could be a good assistant tool to research the process of double glow plasma surface alloying.