【作者】 宋彦；

【导师】 高慧斌；

【作者基本信息】 中国科学院研究生院（长春光学精密机械与物理研究所） ， 机械电子工程， 2010， 博士

【摘要】 伺服系统的速度平稳度是衡量伺服系统性能的重要指标之一,光电测控仪器、精密测试转台等设备都对这一指标有明确要求。当伺服系统工作速度较低时,由于传感器的分辨率限制以及机械传动环节存在的各种扰动量,影响了伺服系统的速度平稳度。提高伺服系统速度平稳度的关键技术在于:提高传动机构设计、制造和装调水平;提高速度信号的检测精度;优化电机驱动控制技术;改进伺服控制技术,增强控制系统的抗扰能力。本文从自动控制技术的角度出发,围绕相关关键技术,展开论文研究,主要研究内容和成果如下:1.采用基于伺服系统内部动态的数学模型描述系统,采用机理分析和实验验证相结合的方法,建立了扰动环节的数学模型。2.分析了采用编码器位置信号做差分测速的不足,提出了采用基于系统动态信息的状态观测器测速方法和基于运动学信息的Kalman滤波测速方法。实验证明,在给定速度为1°/s时,这两种方法较差分测速精度提高了约30%;辅以Newton预测法,较好的解决了带宽和相角滞后间的矛盾。3.分析了摩擦扰动和电机力矩波动影响速度平稳度的模式。通过理论分析,论证了:当给定速度较高时,摩擦扰动对伺服系统的影响是产生一定的稳态误差;当给定速度在Stribeck速度附近时,摩擦扰动影响了系统的稳定性,使速度输出产生了周期震荡的极限环现象。进一步,本文通过对极限环Jacobi矩阵的分析,得出极限环是稳定极限环的结论。最后,通过仿真和实验,验证了上述结论。采用理论分析和实验验证的方法,证明了电机力矩波动可以视为伺服系统受到正弦形式的扰动,对伺服系统的影响是使速度产生了周期波动。这一部分的研究内容旨在揭示主要扰动量对伺服系统速度平稳度的影响模式,为后续控制器的设计奠定基础。4.针对电机力矩波动和摩擦扰动设计了相应的补偿控制算法,达到减小扰动量对伺服系统的影响,提高速度平稳度的目的。针对电机的力矩波动,设计了基于鲁棒自适应控制的扰动力矩补偿算法。该算法采用最小二乘法辨识相关参数,通过数学模型计算补偿量,并在控制量输出中加以体现。仿真结果表明:与超前—滞后+积分控制相比,速度波动由0.5805%(RMS)降低到0.002%(RMS)。同时,通过仿真验证了系统存在白噪声干扰和未建模动态特性时,其误差峰—峰值仍在5%以内,说明该控制算法仍然能够保持原有性能。对于摩擦扰动,采用了一种基于滑模自适应控制的补偿算法。该算法在对摩擦扰动进行自适应补偿的同时,还利用滑模控制的强鲁棒性来保证系统对一些建模误差和不确定性的鲁棒性。仿真结果表明:在等速运动时,速度稳态误差由采用超前—滞后校正的30.95%下降到几乎为0;在给定正弦信号时,速度误差由采用超前—滞后校正的35%(RMS)下降到13%(RMS),该算法能够有效抑制摩擦扰动对速度稳定度的影响;当存在系统参数时变以及存在部分未建模信息的情况下,等速运动的速度误差同样在0附近,正弦运动的误差为19%(RMS),说明该控制策略仍然能够保持对摩擦扰动的有效抑制。5.在论文的最后部分,从工程应用的角度出发,首先提出了对超前—滞后控制策略的一些改进措施,以提高系统的抗扰性能。实验证明,加入积分控制和采用加速度反馈控制时将速度波动由2.92%分别降低到0.9%和0.47%(RMS),由对实验结果的分析得出了以下结论:力矩波动是影响本实验转台速度稳定度的主要因素。随后,采用对力矩波动的自适应补偿控制做了实验验证,实验结果表明,将速度波动进一步降低到0.39% (RMS)。更多还原

【Abstract】 The stability of velocity is one of key indexes of servo system, e.g. opti-cal-electrical tracking system, test turntable call for high level of this index. Espe-cially when the servo system working in the low velocity, servo system face to the problems such as limitation of resolution and kinds of disturbance of transmission. So, the rate of velocity stability degrades. This paper mainly focus to improve the stability of velocity. The context and main conclusion are in the below:The mathematical model is established based on the dynamic characteristic. Mechanism analysis and experiment result are used to establish mathematical model of disturbance.The source of low velocity estimation errors and the shortcomings of the single period difference were analyzed. The state observer and Kalman filter are proposed to estimate the velocity. These methods are based on the dynamic and kinematics char-acteristic of servo system. Experiment results show that, when the input is 1°/s, these methods reduce the estimation error nearly 30%. With the method of Newton predic-tor, the conflict between bandwidth and phase lag are solved.The pattern of how friction and force ripple influence the stability of velocity is analyzed. Friction may cause the servo system has steady state error. when the servo system run near Stribeck velocity, the friction may induce limit cycle. Because the friction influences the stability of closed loop system. Force ripple may cause the ve-locity of servo system has some cycle fluctuation. The adaptive robust control strategy is used to design the compensation to the disturbance. As to the force ripple, the RLS method is used to identify the certain parameters of disturbance. The compensation part is calculated based on the parameter and math model. As to the friction, beside compensate friction on line, the sliding model control is used to improve the robust-ness to some unmodeled dynamic and fluctuation of parameter.At last, from the aspect of engineering application, some methods are verified in turntable. These methods include: the improved lead-lag with the integrator, the lead-lag control with acceleration and integrator, the adaptive robust control with force ripple compensation. Compared to lead-lag method which velocity fluctuation is 2.92%(RMS), the first method reduce to 0.9%, the second method reduce to 0.47%, the third method reduce to 0.39%.更多还原