基于模糊神经网络的多股簧钢丝张力控制系统  

The Wire Tension Control System of Stranded Wire Helical Spring Based on Fuzzy Neural Network

作　　者：张其[1] 王时龙[1] 杨文翰 蔡万强 ZHANG Qi;WANG Shi-long;YANG Wen-han;CAI Wan-qiang(The State Key Laboratory of Mechanical Transmission,Chongqing University,Chongqing 400044,China)

机构地区：[1]重庆大学机械传动国家重点实验室,重庆400044

出　　处：《机械设计与制造》2018年第12期105-108,共4页Machinery Design & Manufacture

基　　金：国家自然科学基金(51375508);教育部创新团队发展计划项目(IRT_15R64)

摘　　要：张力控制是多股螺旋弹簧加工过程中的关键技术,它的控制精度直接影响产品的合格率。为降低控制系统的复杂性,采用并列的单输入单输出控制结构。基于对加工过程中张力波动来源的分析,针对系统的时变性、非线性和钢丝间的相互影响等特点,在模糊神经网络中引入误差与总偏差之比的模糊推理来修正网络的输出值,实现了同时对各股钢丝张力的自适应在线控制。通过试验分析,在加工过程中钢丝张力的波动标准差约为5N,调整时间约为3s,验证了控制系统的快速性和稳定性,较好地解决了多股簧加工过程中钢丝张力一致性的难题。Tension control is the key technology in the winding process of stranded wire helical spring. And its control precision directly affects the qualified rate of products. To reduce the complexity of the control system, parallel single-input and single-output control structures are adopted. The system is time-varying arid nonlinear, furthermore, it contains interaetions between wires. To achieve the adaptive online tension control of wires, based on the analysis of the sourees of tension fuctuation, the fuzzy reasoning of the ratio （between error and total deviation） is introduced to modify the output value of fuzzy neural network. The starzdard deviation of the tension fluctuation is about 5N in the winding process and the setting time is nearly 3s. The experimental results prove that the control system has fast response speed and good stability. It suecessfully solves the problem of tension control of wires during the winding process of strcazded wire helical spring.

关 键 词：多股螺旋弹簧 张力控制 时变 非线性 模糊神经网络 

分 类 号：TH16[机械工程—机械制造及自动化] TP273[自动化与计算机技术—检测技术与自动化装置]