伺服系统在线参数自整定及优化技术研究
发布时间：2018-09-10 09:02
【摘要】：整定伺服系统PI调节器必然要考虑系统等效的转动惯量，同时PI调节器的比例和积分增益主要决定伺服系统的稳定性和响应速度。在线伺服控制器整定技术一般包括转动惯量辨识过程和PI参数整定过程，而且在辨识和整定过程中应避免干预伺服系统的正常工作指令。本文的完全在线参数自整定研究是基于3.3kW的永磁同步电机伺服平台。 首先本文将介绍两种在线转动惯量辨识算法，并针对含有协方差矩阵的辨识算法存在计算字长要求高、在嵌入式处理器多次迭代会累积截断误差、加重噪声对辨识结果影响的问题，提出能削弱此影响的基于遗忘因子递推平方根的在线惯量辨识算法。实验结果表明该方法确实有效，基本不受电机转速给定的影响，辨识精度稳定。同时本文提出若干改进措施，使辨识算法适应负载转矩偶尔变化的场合。 其次，提出一组包含伺服系统速度环开环截止频率、相角裕度和转动惯量的PI整定公式。大量仿真结果说明此PI整定公式具有整定相角裕度和闭环带宽的可行性。随后本文通过大量仿真确定出相角裕度60°下伺服系统速度环极限带宽，，并给出极限带宽通用计算公式。 最后，本文给出PI整定公式在标幺化数字控制系统中的推导方法；同时将FFRSR在线转动惯量辨识法、已提出的PI整定公式和极限带宽计算公式三部分结合形成无需人为干预的完全在线参数自整定模块。实验结果表明整定过程完全没有人工干预，整定效果稳定良好，改善了伺服系统动态响应性能。
[Abstract]:The tuning servo system PI regulator must consider the equivalent moment of inertia of the system, and the proportion and integral gain of the PI regulator mainly determine the stability and response speed of the servo system. On-line servo controller tuning techniques generally include the process of inertia identification and PI parameter tuning, and the interference of normal working instructions of servo system should be avoided in the process of identification and tuning. In this paper, the complete online parameter self-tuning is a permanent magnet synchronous motor servo platform based on 3.3kW. First of all, this paper introduces two online moment of inertia identification algorithms. For the identification algorithm with covariance matrix, the computing word length is very high, and the truncation error is accumulated in the embedded processor. This paper presents an on-line inertia identification algorithm based on the recursive square root of forgetting factor, which can weaken the influence of noise on the identification results. The experimental results show that the method is effective and the identification accuracy is stable. At the same time, some improvement measures are proposed to adapt the identification algorithm to the occasion of occasional load torque variation. Secondly, a set of PI tuning formulas including the open loop cutoff frequency, phase angle margin and moment of inertia of the speed loop of the servo system are proposed. A large number of simulation results show that the PI tuning formula has the feasibility of tuning phase angle margin and closed-loop bandwidth. Then through a lot of simulations, the limit bandwidth of speed cycle of servo system is determined at 60 掳phase angle margin, and the general calculation formula of the limit bandwidth is given. Finally, the derivation method of PI tuning formula in the standard unitary digital control system is given, and the FFRSR on-line moment of inertia identification method is also given. The proposed PI tuning formula and the limit bandwidth calculation formula are combined to form a complete on-line parameter self-tuning module without human intervention. The experimental results show that there is no manual intervention in the tuning process, the tuning effect is stable and the dynamic response performance of the servo system is improved.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM921.541
本文编号：2233991
[Abstract]:The tuning servo system PI regulator must consider the equivalent moment of inertia of the system, and the proportion and integral gain of the PI regulator mainly determine the stability and response speed of the servo system. On-line servo controller tuning techniques generally include the process of inertia identification and PI parameter tuning, and the interference of normal working instructions of servo system should be avoided in the process of identification and tuning. In this paper, the complete online parameter self-tuning is a permanent magnet synchronous motor servo platform based on 3.3kW. First of all, this paper introduces two online moment of inertia identification algorithms. For the identification algorithm with covariance matrix, the computing word length is very high, and the truncation error is accumulated in the embedded processor. This paper presents an on-line inertia identification algorithm based on the recursive square root of forgetting factor, which can weaken the influence of noise on the identification results. The experimental results show that the method is effective and the identification accuracy is stable. At the same time, some improvement measures are proposed to adapt the identification algorithm to the occasion of occasional load torque variation. Secondly, a set of PI tuning formulas including the open loop cutoff frequency, phase angle margin and moment of inertia of the speed loop of the servo system are proposed. A large number of simulation results show that the PI tuning formula has the feasibility of tuning phase angle margin and closed-loop bandwidth. Then through a lot of simulations, the limit bandwidth of speed cycle of servo system is determined at 60 掳phase angle margin, and the general calculation formula of the limit bandwidth is given. Finally, the derivation method of PI tuning formula in the standard unitary digital control system is given, and the FFRSR on-line moment of inertia identification method is also given. The proposed PI tuning formula and the limit bandwidth calculation formula are combined to form a complete on-line parameter self-tuning module without human intervention. The experimental results show that there is no manual intervention in the tuning process, the tuning effect is stable and the dynamic response performance of the servo system is improved.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM921.541
【参考文献】
相关博士学位论文 前1条
1 纪科辉;低速交流电机伺服系统的研究与实现[D];浙江大学;2013年
本文编号：2233991
本文链接：https://www.wllwen.com/kejilunwen/dianlilw/2233991.html
