感应电机模糊自适应全阶磁链观测器和转速估计方法研究

发布时间：2018-08-30 12:23

【摘要】：感应电机由于其具有体积小、重量轻、价格低、转动惯量小和维护简单等优点,在工业中得到了广泛的应用。转速闭环是高性能交流调速系统中必不可少的环节。然而,转速传感器的安装会带来诸如成本增加、可靠性降低、应用场合受限等负面影响。因此,无速度传感器控制技术应运而生,并成为了交流调速领域的研究热点之一。本文围绕感应电机无速度传感器矢量控制系统中磁链观测和转速估计两项关键技术问题展开深入研究。首先,研究转子磁链观测方法。在按照转子磁链定向的矢量控制系统中,转子磁链的准确观测是保证矢量控制系统转子磁链定向成功的必要条件。介绍了基于电流模型和电压模型的转子磁链观测的方法,分析了电机参数对转子磁链观测器影响,分别推导了两种方法的观测转子磁链与实际转子磁链的比值函数,基于此分析了两种模型对电机参数的敏感性。为了在全速范围内都能实现转子磁链的准确观测,结合两种模型的优势,推导出电压电流组合模型磁链观测器,通过一个PI过渡环节,实现电流模型和电压模型的高低速平滑切换。采用梯形法完成对电压电流组合模型磁链观测器的离散化,为构建数字化的感应电机无速度矢量控制系统奠定基础。其次,研究转速估计方法。针对基于全阶状态观测器的感应电机矢量控制系统存在低速不稳定问题,在观测转子磁链定向坐标系下,将转速辨识系统的稳定性问题转换为系统零极点的稳定性问题。利用劳斯判据得到了系统低速发电制动运行时的不稳定区域,给出了一种使不稳定区域最小化的反馈增益矩阵的设计方法。通过分析转速辨识对定、转子电阻的敏感性,提出一种引入自修正因子的模糊转速自适应律设计方法以减小参数变化的影响,提高转速辨识的精度。然后,为了进一步实现鲁棒性更强的无速度传感器矢量控制系统,提出T-S模糊状态观测器。考虑感应电机的运动方程,利用扇区化分的方法,构建五阶感应电机T-S模糊状态观测器。由于反馈增益矩阵是保证T-S模糊状态观测器稳定的关键,采用Lyapunov理论和D-稳定性分析方法,将观测器的稳定性转化为一系列的线性矩阵不等式(LMI),通过Matlab中LMI工具箱求得每一个子系统的反馈增益矩阵。最后,基于Matlab仿真平台和TMS320F28055电机控制实验平台,分别对基于电压电流组合模型磁链观测器、模糊自适应全阶磁链观测器、T-S模糊状态观测器的感应电机无速度传感器矢量控制系统做了大量的仿真和实验研究。验证了本文所提方法的正确性和有效性。
[Abstract]:Induction motor has been widely used in industry because of its advantages such as small volume, light weight, low price, small moment of inertia and simple maintenance. Speed closed loop is an essential link in high performance AC speed regulation system. However, the installation of rotating speed sensor will bring negative effects such as increased cost, lower reliability and limited application. Therefore, the speed sensorless control technology emerges as the times require, and has become one of the research hotspots in the field of AC speed regulation. This paper focuses on two key technologies of flux observation and speed estimation in sensorless vector control system of induction motor. Firstly, the rotor flux observation method is studied. In the vector control system based on the rotor flux orientation, the accurate observation of the rotor flux is the necessary condition to ensure the rotor flux orientation of the vector control system. The method of rotor flux observation based on current model and voltage model is introduced. The effect of motor parameters on rotor flux observer is analyzed. Based on this, the sensitivity of the two models to motor parameters is analyzed. In order to realize the accurate observation of rotor flux in the full speed range, combined with the advantages of the two models, the combined voltage and current model flux observer is derived, and a PI transition link is adopted. The high and low speed smooth switching between current model and voltage model is realized. The trapezoidal method is used to discretize the flux observer of voltage and current combination model, which lays a foundation for the construction of digital speed vector control system for induction motor. Secondly, the method of speed estimation is studied. Aiming at the problem of low speed instability in the vector control system of induction motor based on full order state observer, the stability problem of speed identification system is transformed into the stability problem of zero pole point in the observed rotor flux oriented coordinate system. The unstable region of the low speed power generation brake is obtained by using the Rouse criterion, and a design method of feedback gain matrix to minimize the unstable region is given. By analyzing the sensitivity of rotor speed identification to stator and rotor resistance, a design method of fuzzy speed adaptive law with self-correction factor is proposed to reduce the influence of parameter change and improve the accuracy of speed identification. Then, in order to achieve a more robust speed sensorless vector control system, T-S fuzzy state observer is proposed. Considering the motion equation of induction motor, the fifth order T-S fuzzy state observer of induction motor is constructed by sectioning method. Because the feedback gain matrix is the key to the stability of T-S fuzzy state observer, Lyapunov theory and D- stability analysis method are used. The stability of the observer is transformed into a series of linear matrix inequalities (LMI),). The feedback gain matrix of each subsystem is obtained by LMI toolbox in Matlab. Finally, based on Matlab simulation platform and TMS320F28055 motor control experiment platform, the flux observer based on voltage and current combination model is proposed. The speed sensorless vector control system of induction motor based on fuzzy adaptive full order flux observer / T-S fuzzy state observer has been studied by simulation and experiments. The correctness and validity of the proposed method are verified.
【学位授予单位】：西南石油大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM346;TP273

【参考文献】