首页|期刊导航|机械与电子|参数摄动下耦合电机非奇异滑模并联补偿控制

参数摄动下耦合电机非奇异滑模并联补偿控制OA

Nonsingular Sliding Mode-based Parallel Compensation Control for Coupled Motors Systems Under Parameter Perturbations

中文摘要英文摘要

针对双惯量弹性耦合驱动系统中参数摄动与外部扰动造成负载端转角跟踪精度退化的问题,提出非奇异终端滑模并联补偿策略.在原有反步滑模神经网络控制器之外独立增设补偿支路,构造饱和切换型附加力矩,利用其有限时间收敛特性削减稳态偏差,同时神经网络继续在线逼近未建模动态.采用Lyapunov 推导闭环误差一致最终有界的充分条件,并给出补偿增益整定准则.在 MATLAB/Simulink 中经单位阶跃、慢变扰动叠加快变参考及快变扰动叠加快变参考 3 类工况验证.结果表明,负载端角位置峰值跟踪误差较原方法分别降低 77.4%、39.0%和 32.2%,输出力矩抖振未明显增加.该策略为参数不确定与干扰并存下的高精度伺服控制提供了可行参考.

To address the problem of degraded load-side angular position tracking accuracy in dual-inertia elastically coupled drive systems caused by parameter perturbations and external disturbances,a nonsingular terminal sliding-mode parallel compensation strategy is proposed.Apart from the original backstepping sliding-mode neural network controller,an independent compensation branch is additionally incorporated to construct a saturation-switching type additional torque.Leveraging its finite-time con-vergence property,the steady-state deviation is effectively reduced,while the neural network continues to online approximate the unmodeled dynamics.Sufficient conditions for uniform ultimate boundedness of the closed-loop errors are derived via Lyapunov analysis,and a tuning criterion for the compensation gain is provided.The proposed scheme is validated in MATLAB/Simulink under three test conditions:unit step response,slow-varying disturbance superimposed on fast-varying reference,and fast-varying disturb-ance superimposed on fast-varying reference.Results show that the peak tracking errors of the load-side angular position are reduced by 77.4%,39.0%,and 32.2%,respectively,compared with the original meth-od,without appreciable increase in output torque chattering.This proposed approach provides a practical reference for high-precision servo control under coexisting parameter uncertainties and disturbances.

杨晓武;陈昌云;任洪飞

贵州理工学院人工智能与电气工程学院,贵州 贵阳 550025||贵州理工学院贵州省新型电力系统运行控制重点实验室,贵州 贵阳 550025贵州理工学院人工智能与电气工程学院,贵州 贵阳 550025贵州理工学院人工智能与电气工程学院,贵州 贵阳 550025

信息技术与安全科学

耦合电机反步滑模神经网络非奇异终端滑模

coupled motorsbackstepping sliding modeneural networksnonsingular terminal sliding mode

《机械与电子》 2026 (9)

17-25,9

贵州理工学院高层次人才科研启动经费项目(2023GCCZK)贵州省新型电力系统运行控制全省重点实验室子课题(ZSYS(2025)007)贵州理工学院2021年度学术新苗培养及创新探索专项项目培育项目(GZLGXM-07)

评论