材料表面激光强化类别控制的BP神经网络模型的建立

通过试验分析，表明使用不同的激光工作参数，对材料进行激光强化处理，所得材料表面可归为四种类别，即：未相变硬化，相变硬化，表面微熔及表面熔凝。本文建立了激光工艺参数与材料表面强化类之间关系的BP神经网络模型，并应用该模型，对常用于制造农业机械和发动机齿轮、凸轮轴、链轮、曲轴等零件的材料HT300进行激光强化处理试验。结果表明，BP神经网络模型町方便、准确地选择激光丁艺参数，控制材料表面强化类别及下作性能。

Artificial neural network establish on controlling method for different classifications of laser surface strengthening process

Experiments show that metal will be more or less modified their surface properties by laser strengthening treatment.In this paper four different strengthening classification of structure and characteristic of phase layer:non-transformation hardening,transformation hardening,shallow melting and melting are analyzed and the relationship between the four strengthening classification and laser processing parameters:laser power(P),laser processing beam diameter(D),laser scanning velocity(V) are established by using BP neural network.HT300,as a main high strength cast iron,widely used as gears,camshafts,chain wheel et al.The study results,used HT300 as experimental material,show that laser processing parameters can be chosen conveniently and material surface quality is controlled effectively.