【作者】 徐占国；

【导师】 邵诚；

【作者基本信息】 大连理工大学 ， 控制理论与控制工程， 2010， 博士

【摘要】 电动汽车在节能和环保方面具有较大优势,已经成为近年来发展最快的一种新型汽车。电机及其电驱动系统作为电动汽车的重要组成部分,其工作性能好坏直接影响着整车的各项运行指标。感应电机具有结构坚固、体积小、免维护、价格便宜,能用于易燃易爆环境等优点,而且随着矢量控制性能的不断提高,感应电机已经成为电动汽车的主流驱动电机。电动车用感应电机矢量控制系统与一般的工业应用系统不同,运行效率最大化、调速范围宽、高可靠性和较好的参数鲁棒性是其控制重点和难点。本文围绕这些问题展开研究,取得了以下几方面的研究成果：针对感应电机最大效率控制时的损耗模型的建模问题,从分析定子铁损耗和转子铁损耗与转差率及同步角频率的关系入手,建立了同时考虑定转子漏感、定子铁损耗和转子铁损耗时的感应电机损耗模型。基于该模型,利用拉格朗日优化算法,给出了感应电机最大效率运行时励磁电流与当前负载和转速的定量关系式。并进一步获得了转子相对铁损耗与同步角频率的关系,从理论上说明了在低速区转子铁损耗对电机运行效率有着显著影响。实验结果表明所建损耗模型在全速范围内都有着较高的准确性。为了提高感应电机最大效率控制时对电机参数的鲁棒性、以及确保效率寻优的稳定性、快速性和准确性,设计了一种新的效率模糊控制器。根据所建立的电机损耗模型计算转子磁链搜索初值,通过设计合理的模糊控制规则实现搜索步长的自适应改变,进而通过不断调整转子磁链来搜索逆变器直流侧输入功率最小值,最终实现效率最大化。同时,采用前馈补偿方法,引入一阶微分环节解决了效率优化过程中的低频转矩脉动和转矩快速响应问题。实验结果验证了所设计的效率模糊控制器的有效性,对于提高电动汽车在一次充电后的续驰里程具有重要的实际意义。针对无速度传感器矢量控制技术在低速时转子磁链定向角和转子速度估计难度大、调速性能差,因此无法满足电动汽车对宽调速范围尤其是低速爬行要求的问题,首先通过将高频信号注入法和传统的改进电压模型结合设计了一种新的转子磁链定向角估计混合模型,解决了转子磁链定向角在全速范围内的准确估计问题。通过使高频信号幅度随速度升高而逐渐减小,可实现在高低速之间不同估计方法的平滑切换。进一步将该模型作为参考模型,将传统的电流模型作为可调模型构建模型参考自适应系统(model reference adaptive system, MRAS),解决了包括低速情况下的转子速度估计问题。为了提高速度估计的准确性,还提出了转子电阻在线识别方法,以实时更新可调模型中的转子电阻。实验结果表明,所提出的转子磁链定向角和转子速度估计方法在从低速到高速的全范围内都是有效的。针对感应电机最大效率控制时励磁互感会因磁路饱和程度不同而变化的问题,提出一种基于MRAS的励磁互感在线辨识新方法。该方法采用一种以各相定子电流和积分后的各相定子电压为变量的特殊形式函数,在两种不同的参考坐标系下,分别作为MRAS的参考模型和可调模型；进而通过求解由稳态电磁转矩方程、转子电压方程和转子磁链方程组成的方程组,实现对感应电机励磁互感的在线辨识。理论分析表明,所提出的辨识方法与转子磁链定向角、逆变器死区时间、以及定转子电阻等电机参数均无关,因此有着较强的鲁棒性和较高的辨识精度。实验结果验证了该辨识方法的有效性。更多还原

【Abstract】 Owing to its advantage in energy savings and environmental protection, as a new genre of automobiles, the Electric Vehicle (EV) has been undergoing rapid development in recent years. As essential components of EV, the performance of the motor and its drive system directly affects the overall performance of the EV. The induction motor has the advantages of durability, reduced size, requiring less maintenance, low cost and its ability in working in explosive and flammable conditions. As the performance of vector control continually improves, the induction motor has become the main drive motor for the EV. Vector control system of induction motor for the EV is different from other industrial systems, in its emphasis and focus on efficiency maximization, wide speed variability, increased reliability and good parameter robustness. These problems are discussed herein and some results are achieved in this paper.Loss model of induction motor under efficiency maximization control is studied. Through analyzing the relationship between the stator/rotor iron loss and slip/synchronization angle frequency, we build a loss model of the induction motor under efficiency maximization control that takes the stator leak inductance, the rotor leak inductance, the stator iron loss and the rotor iron loss into account. Based on the proposed loss model, relationship among the rotor flux, load and the rotor speed under efficiency maximization control is derived using the Lagrange theorem. This relationship henceforth leads to the relationship between the rotor relative iron loss and synchronization angle frequency. So it is shown theoretically that the effect of rotor iron loss on the efficiency of induction motor is distinct at low speed. The experimental results show that the proposed loss model is accurate over the full speed range.A new efficiency controller based on fuzzy logic is designed to improve the robustness of the parameters of the induction motor under efficiency maximization control, and the stability, rapidity and accuracy of the process of efficiency optimization. By assigning the searching initial value of the new efficiency controller, determined by the proposed loss model of the induction motor; and its searching step, adjusted automatically according to fuzzy rule, the DC input power of inverter is minimized by changing the rotor flux, so the efficiency maximization is achieved ultimately. The problems associated with low frequency pulsating torque and torque fast response are addressed by using a feedforward compensation algorithm and a first order differentiator. The experimental results show that the proposed efficiency fuzzy controller is effective and is meaningful to improving the EV running distance after a battery charge.The difficulty to estimate the rotor flux orientation angle and the rotor speed at low speed under speed sensorless vector control, and poor speed control, lead to failure to meet the demand for widely variable speed of EV, especially while running at low speed range. Firstly, a novel rotor flux orientation angle estimation hybrid model integrating high frequency signal injection method (HFSIM) and the modified voltage model (MVM) is designed, so it addresses the problem of the rotor flux orientation angle estimation over a full speed range. The high frequency signal amplitude decreases with an increase of speed, so the smooth transition between the different estimation methods from low to high speed range is realized. Furthermore, by proposing the hybrid model serving as the reference model, and a conventional current model serving as the adjustable model, a model reference adaptive system (MRAS) is established, which addresses the problem of the rotor speed estimation including low speed. The rotor resistance online identification scheme is proposed to update the rotor resistance contained in the adjustable model and to ensure the speed estimation accuracy. The experimental results show that the proposed rotor flux orientation angle and rotor speed estimation methods are effective from low to high speed range.The mutual inductance of induction motor is variable under efficiency maximization control, a new method for online identification of the mutual inductance based on MRAS is proposed. A special function, whose variables are the stator current and the stator voltage integral, serves as the reference model and adjustable model of MRAS separately in different reference frames. Also, the equation group, containing steady-state electric torque equation, rotor voltage equation and rotor flux equation, is established and solved to achieve the online identification of the inductance motor mutual inductance. It is shown in theory that the proposed identification method is independent of the rotor flux orientation angle, the dead time of the inverter, and the motor parameters, including the stator and rotor resistance. Good robustness and high precision are achieved. The experimental results show that the proposed identification method is effective.更多还原