基于超级基神经网络的自适应反演非奇异滑模纱线恒张力控制  

Hyper basis function-based adaptive inverse non-singular method for constant-tension yarn transport

作　　者：王罗俊 彭来湖[1] 熊叙一 李杨 胡旭东[1] WANG Luojun;PENG Laihu;XIONG Xuyi;LI Yang;HU Xudong(Key Laboratory of Modern Textile Machinery&Technology of Zhejiang Province,Zhejiang Sci-Tech University,Hangzhou,Zhejiang 310018,China;School of Automation,Zhejiang Polytechnic University of Mechanical and Electrical Engineering,Hangzhou,Zhejiang 310053,China)

机构地区：[1]浙江理工大学浙江省现代纺织装备技术重点实验室,浙江杭州310018 [2]浙江机电职业技术大学自动化学院,浙江杭州310053

出　　处：《纺织学报》2025年第2期92-99,共8页Journal of Textile Research

基　　金：国家重点研发计划项目(SQ2023YFB3200093);浙江省“高层次人才特殊支持计划”科技创新领军人才项目(2023R5212);浙江机电职业技术大学科教融合重点培育项目(A-0271-24-209);浙江省教育厅一般科研项目(Y202456420)。

摘　　要：为解决针织圆机高速工作时纱线张力波动较大问题,提出了一种基于超级基(HBF)神经网络区间观测器的反演非奇异滑模纱线恒张力控制方法。通过构建运动纱线系统的数学模型,运用神经网络逼近系统参数(输纱器与编织机构转动惯量)变动所导致的不确定性响应,将HBF神经网络与区间观测器相结合设计了一个区间状态观测器,估算出系统转速及纱线张力的边界范围,提高了状态识别的准确性。基于纱线张力估算值,构建反演非奇异终极滑模控制器,确保了张力跟踪误差能够在短时间内迅速收敛,从而增强了系统的鲁棒性与动态响应能力。仿真和实验结果表明:所提控制方法成功地使运动纱线张力在1.6 s内达到并维持在预设值,调节时间相较于标准滑模控制及现有文献中的滑模控制器分别缩短了57%和33%,验证了该控制算法的高效性与可靠性。Objective In high-speed and precision knitting process,the complex dynamic behavior of yarn transmission not only affects the accuracy of tension control,but also increases the complexity and maintenance cost of the system.It is hence necessary to explore new control methods to improve the accuracy and reliability of yarn tension stability control.Sensorless tension control method reduces the dependence on sensors.By optimizing the structure and material of the yarn transmission mechanism,the influence of adverse factors such as vibration and friction is reduced,and the production efficiency and product quality of the circular machine are improved.Method The yarn motion during knitting was decoupled into two independent systems using the inversion method,and an inverse non-singular terminal sliding mode controller was designed to improve the sliding mode surface to make the yarn real-time tension error converge quickly in a short time.The hyper basis function(HBF)neural network was introduced into the interval state observer of the yarn transmission system,which was close to the random response caused by the changes of parameters such as weft storage radius and the inertia of the knitting area.Results The designed HBF neural network interval observer was used to estimate the boundary value of the moving yarn system.After the operation of the three controllers,the controller designed in this research was shown to stabilize the tension in 1.6,which is significantly better than the 3.5 s of the conventional sliding mode and the 2.4 s described in related literature,and the adjustment time was reduced by 57%and 33%respectively.The experimental results showed that the sliding mode controller designed in this paper has faster response and higher tracking accuracy,which is significantly better than the other two controllers.The traditional proportional-integral-differential(PID)controller performed the worst for the yarn relaxation problem when the moving yarn system is started,while the improved sliding mode controller

关 键 词：纱线张力 超级基神经网络 状态观测器 张力误差 滑模控制器 针织圆机 

分 类 号：TS181.8[轻工技术与工程—纺织材料与纺织品设计] TS103.7[轻工技术与工程—纺织科学与工程]