异步电机电流模型预测控制技术研究

发布时间：2018-10-05 18:22

【摘要】：模型预测控制算法稳定性强,思想简单,动态响应快,容易处理非线性约束,被认为是继矢量控制和直接转矩控制之后电机控制领域的新一代高性能控制策略之一。本文在磁场定向控制中引入模型预测控制理论,对异步电机电流模型预测控制的控制原理、控制性能进行了较为深入的研究,并利用仿真以及实验验证了控制策略的可行性和有效性。然后,基于模型预测的基本原理给出了预测控制的三个基本步骤;基于异步电机数学模型和状态方程建立旋转坐标系下的电流预测模型;基于磁场定向策略构建电流模型预测控制的基本框架;针对异步电机转子磁链观测,在电压型转子磁链观测器中引入基于离散滤波器的改进优化措施,通过设置合适的截止频率与补偿措施,解决了积分饱和以及直流偏置问题,提高了预测控制磁链观测的精度。此外,面向单步电流模型预测控制的控制过程,利用对电流未来轨迹线性拟合的方式以解决单步电流模型预测控制算法可能造成的发散、震荡问题;针对电流预测控制中存在的一拍延时问题,提出了一拍延时补偿策略,改善了一拍延时造成的不利影响;针对预测控制下平均开关频率较高的问题,引入双边界圆限制策略以减小开关动作次数;针对低采样频率下电流预测控制中电流误差较大问题,引入误差预测策略,以将电流误差维持在误差限以下;针对预测控制采样频率高的问题,引入了双矢量电流模型预测控制的思想。为验证电流模型预测控制的理论分析效果,搭建了 MATLAB模型进行仿真验证,仿真评估了电流模型预测控制下电机的动态响应速度、验证了电流模型预测理论基础的正确性、检验了误差预测策略与一拍延时补偿算法的控制性能、对比了单矢量预测控制与双矢量预测控制的控制效果。同时,文中搭建了异步电机电流模型预测控制实验系统的软硬件平台,实验结果也验证了本文提出的电流模型预测控制策略的有效性。
[Abstract]:The model predictive control (MPC) algorithm is considered to be one of the new generation of high performance control strategies in the field of motor control after vector control and direct torque control because of its strong stability, simple thinking, fast dynamic response and easy handling of nonlinear constraints. In this paper, the theory of model predictive control is introduced in the field oriented control, and the control principle and control performance of current model predictive control for asynchronous motor are studied deeply. The feasibility and effectiveness of the control strategy are verified by simulation and experiments. Then, based on the basic principle of model prediction, three basic steps of predictive control are given, and the current prediction model in rotating coordinate system is established based on mathematical model and state equation of asynchronous motor. The basic framework of current model predictive control based on flux orientation strategy is constructed, and an improved optimization method based on discrete filter is introduced to the rotor flux observer for induction motor. By setting appropriate cutoff frequency and compensation measures, the problems of integral saturation and DC bias are solved, and the precision of flux observation in predictive control is improved. In addition, for the control process of single-step current model predictive control, the problem of divergence and oscillation caused by single-step current model predictive control algorithm is solved by linear fitting of the future current trajectory. Aiming at the problem of one-beat delay in current predictive control, the compensation strategy of one-beat delay is put forward to improve the adverse effect caused by one-beat delay, and the problem of high average switching frequency under predictive control is also proposed. Double boundary circle limiting strategy is introduced to reduce the number of switching operations, and the error prediction strategy is introduced to maintain the current error below the error limit in the current predictive control at low sampling frequency. To solve the problem of high sampling frequency of predictive control, the idea of double vector current model predictive control is introduced. In order to verify the theoretical analysis effect of current model predictive control, the MATLAB model is built to simulate and verify the dynamic response speed of the motor under the current model predictive control, and the correctness of the theoretical basis of current model prediction is verified. The control performance of error prediction strategy and one-beat delay compensation algorithm is tested, and the control effects of single-vector predictive control and double-vector predictive control are compared. At the same time, the hardware and software platform of the current model predictive control experiment system for asynchronous motor is built. The experimental results also verify the effectiveness of the proposed current model predictive control strategy.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM343

【参考文献】