基于模糊对向网络的自学习控制器及其在喷焊中的应用

Fuzzy CP Network-Based Self-Learning Controller and Its Application to Powder Overlaying Welding

为了解决用机械喷焊时熔池温度难以控制的问题 ,将模糊对向网络与自学习模糊控制的理论相结合 ,研制出了一种自学习模糊对向网络控制器 .通过自学习算法 ,控制器能获得足够的相关信息并能不断地优化控制性能 ,适用于非线性时变过程的控制 .喷焊机工作时 ,用该控制器控制熔池温度 ,使熔池温度保持在设定的温度范围 ,获得了高质量的喷焊层 .喷焊工件 30 0mm长度 ,喷焊层厚度误差小于 0 .1mm .

In order to solve the problem that the temperature of melten pool can not be controlled during. Powder overlaying welding, a self_learning controller based on fuzzy CP network is developed by combining fuzzy CP network and the theory of self_learning fuzzy control. With self_learning algorithm, the controller can obtain sufficient information and optimize its control characteristics. The controller can be applied to control of nonlinear time_variational uncertain process. When Powder overlaying welding maching is operating, the controller can control the melten pool in the set region of temperature, and a good quality layer is formed. The error of thickness of layer of 300mm in length is: Δδ≤0.1mm.