基于精确离散化预测模型及迭代学习控制的负载模拟系统转矩波动抑制研究

Permanent magnet synchronous motor under finite control set model predictive direct torque control achieves excellent dynamic and static load simulation performance. It is an ideal new choice for load simulator of traction drive systems considering the change of wheel-rail adhesion state. However, the discretization error of the prediction model and the measurement error of the motor currents will cause the torque ripple. Based on an exact discrete-time predictive model and iterative learning control, the torque ripple minimization of permanent magnet synchronous load motor is studied in this project. First of all, the load characteristics of traction motor is established using Oldrich Polach adhesion formula. The load torque reference under different adhesion can be obtained. Secondly, an accurate discretization method based on the Cayley-Hamilton theorem is proposed. It can avoid the discretization error, which is caused by Taylor truncation in traditional Euler method. Then, considering the repetitive operation characteristics of load simulator, the periodic torque ripple is suppressed using iterative learning control. And its convergence is proved under the condition of finite control set. This project will provide theoretical support for the traction motor load simulator, which is of great significance to promote the innovative design of traction drive system.

有限控制集合模型预测直接转矩控制下的永磁同步电机具有高动态、高精度的加载性能，是轨道车辆牵引传动系统在考虑轮轨黏着状态变化时负载特性模拟的理想选择。然而，预测模型离散化误差和电流测量误差会引起加载电机的转矩波动。本项目将基于精确离散化预测模型以及迭代学习控制，抑制轨道车辆负载特性模拟中的转矩波动问题。首先，为了获取不同黏着状态下加载力矩给定值的计算方法，基于Oldrich Polach黏着力公式建立车辆牵引传动系统负载特性数学模型；其次，为了避免欧拉离散法进行泰勒截断所带来的离散化误差，研究预测模型的Cayley-Hamilton精确离散化方法，以提高加载精度；然后，考虑负载模拟系统重复运行的特点，引入迭代学习控制抑制电流测量误差所引起的转矩波动，并分析其与有限控制集合模型预测控制相结合的收敛性问题。本项目将为牵引传动负载模拟系统提供理论支撑，对促进牵引传动系统的创新设计具有重要意义。

有限控制集合模型预测直接转矩控制下的永磁同步电机具有高动态、高精度的加载性能，是轨道车辆牵引传动系统在考虑轮轨黏着状态变化时负载特性模拟的理想选择。本项目研究了预测模型的精确离散化，提高了预测模型的准确度，一定程度上降低电机的转矩波动；提出了一种序列模型预测直接转矩控制，在不损失动静态性能的前提下实现了无权重系数模型预测控制。分析了定子电流测量误差对转矩波动的影响，研究了有限控制集合迭代学习控制抑制转矩波动，并提出了一种改进型的自适应选择谐波消除法。本项目将为牵引电机负载模拟系统提供理论支撑，对促进牵引传动系统的创新设计具有重要意义。