论文部分内容阅读
计算机数控应力盘光学加工技术有其突出的优势,但应力盘形变却难以有效控制,是实现计算机数控应力盘光学加工中的难题。对此,本文分析了应力盘形变控制系统的特点,引入了CMAC神经网络实现应力盘形变控制的模型。提出了将CMAC神经网络应用于应力盘逆变形智能控制的创意和实现方法,以应力盘面形参数和对应的驱动器电压参数作为样本训练CMAC神经网络,将训练成功的CMAC神经网络作为控制器控制应力盘变形,取得了误差小于5%的计算机仿真结果。
Computer numerical control stress plate optical processing technology has its outstanding advantages, but it is difficult to effectively control the deformation of the stress plate, which is a difficult problem in the optical processing of the computer numerical control stress plate. In this regard, this paper analyzes the characteristics of the stress plate deformation control system, the introduction of CMAC neural network deformation plate deformation control model. The idea and method of applying CMAC neural network to intelligent control of stress plate inverse deformation are proposed. The CMAC neural network is trained by taking the shape parameters of stress plate and the corresponding drive voltage parameters as samples, and the CMAC neural network trained as controller is controlled Stress plate deformation, obtained the error less than 5% of the computer simulation results.