基于参数辨识的永磁同步电机无差拍电流预测控制

发布时间：2018-08-23 13:20

【摘要】：永磁同步电机具有动态响应快、稳态精度高、调速范围宽等优点,在高精度伺服控制领域得到了广泛的应用。伺服系统电流内环的控制效果对电机的动稳态性能影响较大,传统的电流环控制策略难以满足人们对电机控制性能的要求。因此,新的控制策略如预测控制等被用于电机电流环的控制中。本文首先对表贴式永磁同步电机的数学模型进行简化,推导出其在三相静止坐标系下的数学模型,基于磁势守恒原则推导其CLARK和PARK变换矩阵,从而得到三相永磁同步电机在d、q轴下的数学模型,实现电机励磁分量和转矩分量的解耦。通过一阶泰勒公式对其数学模型离散化,得到电机的预测控制模型。其次,对无差拍电流预测控制原理和控制性能进行了研究,由于预测控制的性能依赖于电机参数的准确性,本文对电机参数偏差和预测控制动稳态性能的关系进行了理论分析,当预测模型电感与电机实际电感相差较大时,控制系统发散。为了增强控制系统对电感参数的鲁棒性,采用一种鲁棒电流预测控制算法,通过改变权重系数的大小来调节控制系统的稳定范围。但是权重系数的减小会降低控制系统的带宽,导致电流的动态响应变慢,因此,为了最大程度的提高预测控制的性能,需要结合这两种算法的优点,取长补短。然后,为了消除电机参数偏差导致的电流稳态误差,采用模型参考自适应的方法对电机的电感和磁链进行在线辨识,用电机参数的辨识值去实时的修正预测模型参数。同时,基于电感参数的在线辨识提出了一种在线切换策略,在电感偏差较大时使用鲁棒控制算法提高控制系统的稳定性,而当电感参数收敛到真实值时,再切换回传统无差拍电流预测控制,使得控制系统稳定性提高的同时维持其良好的动态响应。最后,基于TMS320F28335芯片设计了永磁同步电机驱动系统,通过实验进一步研究了无差拍电流预测控制性能和电机参数的关系,验证了鲁棒预测控制算法、模型参考自适应参数辨识算法以及预测控制算法在线切换策略的正确性和有效性。
[Abstract]:Permanent magnet synchronous motor (PMSM) has many advantages, such as fast dynamic response, high steady-state precision and wide speed range, so it has been widely used in the field of high-precision servo control. The control effect of the current inner loop of servo system has a great influence on the dynamic and steady performance of the motor. The traditional current loop control strategy is difficult to meet the requirements of the motor control performance. Therefore, new control strategies such as predictive control are used in motor current loop control. In this paper, the mathematical model of permanent magnet synchronous motor (PMSM) is simplified, and its mathematical model in three-phase stationary coordinate system is deduced. Based on the principle of conservation of magnetic potential, the CLARK and PARK transformation matrices are derived. The mathematical model of the three-phase permanent magnet synchronous motor under dq-axis is obtained, which can decouple the excitation and torque components of the motor. The mathematical model is discretized by the first order Taylor formula and the predictive control model of the motor is obtained. Secondly, the principle and control performance of predictive control are studied. Because the performance of predictive control depends on the accuracy of motor parameters, the relationship between the error of motor parameters and the dynamic and steady performance of predictive control is analyzed theoretically in this paper. The control system diverges when the difference between the predictive model inductance and the actual inductance of the motor is large. In order to enhance the robustness of the control system to the inductance parameters, a robust current predictive control algorithm is used to adjust the stable range of the control system by changing the weight coefficient. However, the decrease of the weight coefficient will reduce the bandwidth of the control system and slow down the dynamic response of the current. Therefore, in order to maximize the performance of predictive control, it is necessary to combine the advantages of the two algorithms to complement each other. Then, in order to eliminate the current steady-state error caused by the parameter deviation of the motor, the inductance and flux chain of the motor are identified online by the model reference adaptive method, and the model parameters are corrected in real time by the identification value of the motor parameters. At the same time, an on-line switching strategy based on inductance parameter identification is proposed. The robust control algorithm is used to improve the stability of the control system when the inductance deviation is large, but when the inductance parameter converges to the real value, the robust control algorithm is used to improve the stability of the control system. Then switching back to the traditional deadbeat current predictive control, the stability of the control system is improved while maintaining its good dynamic response. Finally, a permanent magnet synchronous motor drive system based on TMS320F28335 chip is designed. The relationship between the performance of beat free predictive control and motor parameters is further studied through experiments, and the robust predictive control algorithm is verified. Model reference adaptive parameter identification algorithm and predictive control algorithm are correct and effective.
【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】：2017
【分类号】：TM341

【参考文献】