PMSM参数辨识及其无模型电流预测控制

发布时间：2018-02-21 10:56

本文关键词： 永磁同步电机逆变器非线性补偿 MFAC 参数辨识无模型电流预测控制　出处：《合肥工业大学》2016年硕士论文　论文类型：学位论文

【摘要】：“十三五”规划提出,到2020年建立起完善的电动汽车动力系统科技体系和产业链技术系统,实现各类电动汽车的产业化,促进新能源汽车战略新兴产业进入快速成长期。大力发展电动汽车产业是我国在汽车行业对国外品牌实现弯道超车的重要契机。永磁同步电机(PMSM)以其优越的性能成为电动汽车电驱动系统的主要驱动电机,车用永磁同步电机驱动系统不仅要求调速范围宽,而且要求转矩输出精度高,受逆变器非线性和电机复杂运行工况导致电机参数变化的影响,传统基于PI控制的PMSM驱动系统难以实现转矩的准确控制。此外,逆变器非线性亦会导致PMSM出现输出转矩脉动和定子电流畸变,直接影响PMSM参数在线辨识精度。为此,论文重点研究PMSM参数辨识及其无模型电流预测控制,其研究对于提高PMSM参数辨识精度及PMSM控制器抗参数变化的鲁棒性,无疑具有重要的理论研究价值和工程应用价值。论文首先分析逆变器非线性对逆变器输出电压的影响,并综述了目前主要的补偿方案。在此基础上,建议了基于无模型参考自适应的逆变器非线性补偿方案,并通过系统建模和仿真验证建议方案的可行性和有效性。考虑到永磁同步电机参数辨识对其控制系统设计和电驱动系统运行状况的实时监控非常重要,由于电动汽车的复杂运行工况导致PMSM参数变化范围较大,必需对电机参数实施在线辨识。为此,论文基于现有电机参数辨识方法的梳理分析和总结,给出了基于微分代数的永磁同步电机多参数在线辨识方案,通过系统仿真证实建议的电机参数辨识方案能够解决现有的电机参数辨识方案中存在的多参数在线辨识模型欠秩和辨识精度亟需提升等关键问题。电动汽车永磁同步电机驱动系统通常采用传统的双闭环PI控制,当系统在宽调速范围和参数变化条件下运行时,PI控制易出现饱和失调甚至运行失稳等问题；传统永磁同步电机电流预测控制虽然可以解决PI控制存在的部分不足,但是敏感依赖于电机参数,为此,论文最后研究永磁同步电机的无模型电流预测控制,旨在摆脱PMSM控制对电机参数的依赖且兼顾提升系统的性能,并且通过系统仿真验证其技术优势。
[Abstract]:According to the 13th Five-Year Plan, by 2020, a sound scientific and technological system of electric vehicle power system and a technical system of industrial chain will be established to realize the industrialization of all kinds of electric vehicles. The development of electric vehicle industry is an important opportunity for our country to realize the bend overtaking of foreign brands in the automobile industry. PMSMM (permanent Magnet synchronous Motor) has its superior performance. Become the main driving motor of electric vehicle electric drive system, The drive system of vehicle permanent magnet synchronous motor (PMSM) not only requires a wide range of speed regulation, but also requires a high torque output precision, which is affected by the variation of motor parameters caused by nonlinear inverter and complex operating conditions of motor. The traditional PMSM drive system based on Pi control is difficult to realize the accurate torque control. In addition, the nonlinear inverter will also lead to the output torque ripple and stator current distortion of PMSM, which directly affect the accuracy of on-line identification of PMSM parameters. This paper focuses on PMSM parameter identification and model-free current predictive control, which can improve the accuracy of PMSM parameter identification and the robustness of PMSM controller against parameter change. Undoubtedly, it has important theoretical research value and engineering application value. Firstly, the paper analyzes the influence of inverter nonlinearity on inverter output voltage, and summarizes the main compensation schemes. A nonlinear compensation scheme for inverter based on modelless reference adaptive is proposed. The feasibility and validity of the proposed scheme are verified by system modeling and simulation. Considering that the parameter identification of PMSM is very important to the design of PMSM control system and the real-time monitoring of the operation of the electric drive system, Due to the large range of PMSM parameters due to the complex operating conditions of electric vehicles, it is necessary to carry out on-line identification of motor parameters. A multi-parameter on-line identification scheme for PMSM based on differential algebra is presented. It is proved by system simulation that the proposed scheme of motor parameter identification can solve the key problems such as the lack of rank of multi-parameter on-line identification model and the need to improve the precision of identification of permanent magnet of electric vehicle existing in the existing scheme of motor parameter identification. The traditional double closed loop Pi control is usually used in the drive system of synchronous motor. When the system is running in a wide speed range and variable parameters, the Pi control is prone to saturation misadjustment or even operational instability, while the traditional PMSM current predictive control can solve some shortcomings of Pi control. However, the sensitivity depends on the motor parameters. Therefore, this paper studies the model-free current predictive control of permanent magnet synchronous motor (PMSM) in order to get rid of the dependence of PMSM control on motor parameters and improve the performance of PMSM. The technical advantages are verified by system simulation.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM341;U469.72

【参考文献】