基于速度观测器的PMSM转动惯量辨识研究

发布时间：2018-01-02 04:29

本文关键词：基于速度观测器的PMSM转动惯量辨识研究　出处：《华中科技大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着永磁同步电机(PMSM)交流伺服系统在越来越多的高性能运动控制领域中的广泛应用,如何提高系统的控制精度、稳定性、静态及动态性能就显得尤为重要,然而电机及其负载的转动惯量是影响系统中控制器控制参数的主要因素之一。因此,若能及时辨识出PMSM伺服系统的转动惯量,然后依据其变化对系统中的控制器参数实施调整以补偿其对系统控制带来的影响,就可以提高系统的控制性能。首先,分析了一些常用的参数辨识算法,基于最小二乘法良好的收敛性与无偏性,将递推最小二乘法应用于转动惯量的辨识之中。但是由于递推算法采样次数过大时,新数据会失去对之前的估计值修正的能力,因此本文采用对老数据加上遗忘因数的办法,通过降低老数据的信息量,提高了新数据的影响力,从而获得跟踪参数变化的实时估计。基于此给出了该算法在PMSM伺服系统转动惯量辨识的具体实现方法与形式。其次,由于电机转速通常都是位置传感器获取的位置信号经过简单差分计算得到的。针对因传感器存在固有的相位滞后而影响电机速度计算结果的问题,从提高电机转速计算精度的角度引入了速度观测器。观测器可以降低传感器固有的相位滞后,因此观测到的速度信号相比传感器获得的更精确也更可靠,这对转动惯量辨识的准确度也有一定的提高。最后,采用直轴电流为零的矢量控制策略,构建了PMSM闭环控制系统方案。基于此方案在MATLAB/Simulink中设计了PMSM闭环控制下转动惯量辨识的仿真模型,基于以DSP为核心的VECTODRIVE VD600控制驱动器、登奇永磁同步电动机以及VECTONUM V8工控机的硬件平台搭建了PMSM转动惯量辨识的实验平台,并完成了本平台主要的软件设计。本课题所做的仿真与实验都验证了基于速度观测器与递推最小二乘法相结合的惯量辨识算法的可行性与有效性,实验表明在速度变化的条件下无论是空载还是恒负载,转动惯量辨识的速度与精确度都取得了满意的效果。
[Abstract]:With the wide application of PMSM AC servo system in more and more high performance motion control fields, how to improve the control accuracy and stability of the system. Static and dynamic performance is particularly important, but the moment of inertia of the motor and its load is one of the main factors affecting the controller control parameters in the system. If the moment of inertia of the PMSM servo system can be identified in time, then the controller parameters in the system can be adjusted according to its changes to compensate for its influence on the control of the system. The control performance of the system can be improved. Firstly, some commonly used parameter identification algorithms are analyzed, which are based on the good convergence and unbias of the least square method. The recursive least square method is applied to the identification of the moment of inertia. But when the sampling times of the recursive algorithm are too large, the new data will lose the ability to correct the previous estimates. So this paper adopts the method of adding the forgetting factor to the old data and improves the influence of the new data by reducing the amount of information of the old data. Thus the real-time estimation of tracking parameters is obtained. Based on this, the realization method and form of the algorithm in PMSM servo system moment of inertia identification are given. Secondly. Because the speed of motor is usually obtained by the position signal obtained by the position sensor through simple difference calculation, aiming at the problem that the sensor has inherent phase lag which affects the calculation result of the motor speed. In order to improve the accuracy of motor speed calculation, a speed observer is introduced, which can reduce the inherent phase lag of the sensor, so the observed speed signal is more accurate and reliable than that obtained by the sensor. It also improves the accuracy of the moment of inertia identification. Finally, the vector control strategy with zero direct axis current is adopted. The scheme of PMSM closed-loop control system is constructed. Based on this scheme, the simulation model of moment of inertia identification under PMSM closed-loop control is designed in MATLAB/Simulink. VECTODRIVE VD600 control driver based on DSP. The hardware platform of permanent magnet synchronous motor (PMSM) and VECTONUM V8 industrial control computer is used to set up the experimental platform of PMSM moment of inertia identification. The main software design of this platform has been completed. The simulation and experiment in this paper have verified the feasibility and effectiveness of inertia identification algorithm based on speed observer and recursive least square method. The experimental results show that the speed and accuracy of the moment of inertia identification are satisfactory both under the condition of velocity change and no load or constant load.
【学位授予单位】：华中科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TM341

【相似文献】