超大尺寸高密度碳基复合材料预成型体编织系统多电机协同控制  

Multi-motor cooperative control of braiding system of super-large-size and high-density carbon-based composites

作　　者：张峰[1] 许高平 李硕 季诚昌[1] ZHANG Feng;XU Gaoping;LI Shuo;JI Chengchang(College of Mechanical Engineering,Donghua University,Shanghai 201620,China)

机构地区：[1]东华大学机械工程学院,上海201620

出　　处：《东华大学学报（自然科学版）》2023年第1期76-83,94,共9页Journal of Donghua University(Natural Science)

基　　金：国家重点研发计划资助项目(20017YFB0309800);工信部制造新模式应用资助项目(201746802)。

摘　　要：为提高超大尺寸高密度碳基复合材料预成型体编织系统单元机的经纱开口电机协同性,设计一种粒子群-神经网络速度补偿器和偏差耦合控制结构相结合的控制方式。利用粒子群算法的全局优化能力搜索BP(back propagation)神经网络最佳初始连接权值,克服BP神经网络存在收敛速度慢和容易陷入局部极值的问题。研究结果表明,粒子群-BP神经网络算法模型协同控制能有效提高编织系统经纱开口驱动电机的协同性能,相比传统固定增益控制能更好地实现多伺服电机的同步运行,显著提高编织系统的同步精度和编织效率。In order to improve the synergy of the warp shedding motor of the unit machine of the super-large-size and high-density carbon-based composite preform weaving system, a control method combining a particle swarm-neural network speed compensator and a deviation coupling control structure is designed. Using the global optimization ability of particle swarm algorithm to search for the best initial connection weight of BP(back propagation) neural network, it can overcome the problem of slow convergence speed and easy to fall into local extreme value in BP neural network. The research results show that: the coordinated control of the particle swarm-BP neural network algorithm model can effectively improve the cooperative performance of the warp shedding drive motor of the knitting system, and can achieve the synchronous operation of multiple servo motors compared to the traditional fixed gain control, and significantly improve the synchronization accuracy of the knitting system and weaving efficiency.

关 键 词：超大尺寸三维编织 偏差耦合 粒子群优化 BP神经网络 

分 类 号：TM301.2[电气工程—电机] TM351