体积成型过程的数值仿真与智能控制一体化技术  

Real-Time Self-Adaptive Control of Bulk Forming Process of Forging

作　　者：刘玉红[1] 李付国[1] 吴诗 

机构地区：[1]西北工业大学材料科学与工程学院,陕西西安710072

出　　处：《西北工业大学学报》2003年第1期114-117,共4页Journal of Northwestern Polytechnical University

基　　金：航空基础科学基金 (98H5 3 0 68);西北工业大学博士论文创新基金 (2 0 0 2 12 )资助

摘　　要：从系统观点出发 ,运用控制论原理 ,论证了体积成型过程与电学模型之间具有相似的对应关系。提出了一种体积成型过程的设计与控制一体化技术 ,用有限元数值仿真结果训练人工神经网络 (ANN)来进行体积成型过程的计算 ,由 ANN映射的模拟电路模型与自适应控制系统并联 ,构成模型跟随的自适应控制系统。该模拟电路模型有较强的自适应学习能力 ,并且运算速度快 ,可大大缩短时间延滞 。Quality of forging made from difficult to deform material in P. R. China suffers from the lack of real time control of its bulk forming process. FEM(finite element method) calculation of deformation of forging often requires hours or even tens of hours, thus making impossible the real time control of its bulk forming process, which occurs in just seconds. We bridge this immense time gap by adopting the effective measures described in sections 1 and 2. Subsection 1.3 discusses the analogy which allows the mechanical quantities of the bulk forming process to be converted into electrical ones of a corresponding electrical circuit. Section 2 describes how to use results of FEM numerical simulation to train ANN(artificial neural networks) to perform calculation of the electrical quantities corresponding to the mechanical quantities in the bulk forming process of forging. Section 2 then describes the reconversion of electrical quantities into mechanical quantities as feedbacks to a self adaptive control system for real time control of bulk forming process of forging. Section 3 gives an illustrative example, which shows preliminarily that our proposed method of real time self adaptive control of bulk forming process of forging made from difficult to deform material is indeed feasible.

关 键 词：体积成型 有限元仿真 人工神经网络 模拟电路 模型跟随自适应控制 

分 类 号：TP273.2[自动化与计算机技术—检测技术与自动化装置] TG306[自动化与计算机技术—控制科学与工程]