关键词: 微弧等离子喷涂, 遗传神经网络, 多目标优化

Abstract: According to the test results of the bonding strength and microhardness of nanostructured AT13 coatings prepared by micro-plasma spraying on conditions of different processing parameters, using genetic algorithms and neural networks, the non-linear model on the relationship between the properties of nanostructured AT13 coatings and the process parameters was set up and simulated. The optimization was based on the genetic algorithms and multi-objective optimization theory. The results show that the simulated results agree with experimental values, and the relative error is less than 0.5%. The optimum parameters are that spraying current is 150A, argon flow is 0.64m³/h, argon pressure is 0.38MPa, and the overall properties of nanostructured AT13 coatings are the best.

Key words: micro-plasma spraying, genetic algorithms and neural networks, multi-objective optimization