径向基神经网络在熔融粒子飞行特性预测中的应用  

Application of RBF neural network for forecasting characteristics of in-flight particles by plasma spraying

作　　者：高燕清[1] 方建成[1] 赵紫玉[1] 杨磊[1] 

机构地区：[1]华侨大学机械工程及自动化学院,福建泉州362021

出　　处：《功能材料》2007年第9期1563-1565,共3页Journal of Functional Materials

基　　金：国家自然科学基金资助项目(50675072);福建省自然科学基金资助项目(2006J0165);辽宁省自然科学基金资助项目(20062178)

摘　　要：在等离子熔射成形中,粒子的飞行特性是影响沉积率和涂层动态生长质量的重要因素,而熔射工艺参数又是影响粒子飞行特性的直接因素。以等离子熔射ZrO2粉末为例,采用正交实验的方法,分析了工艺参数与粒子飞行特性间的关系。利用径向基(RBF)神经网络建立了预测模型,实现对熔射过程中飞行粒子温度与速度的预测以及工艺参数的优化。通过对仿真结果与实验结果的比较,表明了该预测模型的有效性。The main factors that influence the deposition efficiency and coating quality are the state of in-flight particles, which are directly influenced by processing parameter during plasma spraying. In this study, plasma spraying of ZrO2 powder was implemented according to the method of orthogonal experiments, and the relationship between processing parameter and characteristics of in-flight particles, which were monitored by an optical monitoring system of CCD camera, were investigated. Radial basis function （RBF） neural network model has been designed to forecast the temperature and velocity of in-flight particles, and optimize processing parameter. The comparison of the simulations with the experimental results shows the validity of the model.

关 键 词：等离子熔射 飞行特性 RBF神经网络 预测模型 

分 类 号：TG664[金属学及工艺—金属切削加工及机床] TK122[动力工程及工程热物理—工程热物理]