大功率三电平PWM整流器优化控制和电网不平衡运行研究

High power three-level rectifier has to solve the conflicting problems of low switching losses, low current harmonics and high dynamic peformance at the same time. At low switching frequency, using synchronous optimal modulation can achieve low current harmonics but the dynamic performance is poor. On the contrary, model predictive control (MPC) is able to achieve quick dynamic response at low switching frequency. Combining the merits of both synchronous optimal modulation and MPC, this project will propose MPC-based synchronous optimal modulation to solve the aforementioned problems for high power rectifiers. Firstly, the MPC based on single control period and its optimization are studied, focusing on the problems of high computational burden and prediction accuracy in multiple-step MPC. Furthermore, multiple-vector-based MPC is proposed to further improve the steady state performance and reduce the switching frequency of single-vector-based MPC. Secondly, the MPC-based closed-loop control of synchronous optimal modulation is studied. By solving the online soltuion to optimal pulsewidth modulation, high performance online closed-loop control based only on MPC is achieved. Finally, the compensating reference value of control variables without the knowledge of line inductance is proposed, which can be integrated into the control methods developed under the condition of balanced grid voltage without modifying the control structure. The control method proposed in this project features the merits of both quick response of MPC and excellent steady state performance of synchronous optimal modulation, and what's more, can be implemented on-line, which provides a novel reliable and safe solution to the optimal control of high power PWM rectifiers.

大功率三电平PWM整流器需要同时解决低开关损耗、低电流谐波和高动态性能几个相互矛盾的问题。在较低的开关频率下，同步优化调制可以获得较小的电流谐波但动态性能差，而模型预测控制(MPC)可以获得快速的动态响应。结合二者优点，本项目采用基于MPC的同步优化脉冲调制来解决上述难题。首先研究基于单个控制周期的MPC及其优化，重点解决多步长MPC中的计算量大和预测精度问题，进一步将单矢量MPC扩展为多矢量MPC以提高稳态性能并降低开关频率。其次研究基于MPC的闭环优化调制技术，在解决同步优化调制在线求解的基础上实现完全基于MPC的在线高性能闭环控制策略。最后，研究不依赖于电感参数的控制补偿指令生成方法，无需改变平衡电网下控制方法的结构而拓展其电网不平衡时的故障穿越能力。本项目提出的控制方法将同时具备MPC的快速响应和同步优化调制的优异稳态性能，且可以在线实现，为大功率整流器优化控制提供了新型解决方案。

项目针对三电平PWM整流器的优化控制及其在电网不平衡下的控制进行了深入研究，提出了一系列原创性成果。针对传统的单矢量DPC应用于PWM整流器时存在的功率脉动大、电流谐波高和开关频率不固定等问题，先后提出了功率脉动最小、功率无差拍和新型简单占空比控制等方法来对DPC选出的矢量时间进行优化，在维持传统DPC动态性能的基础上显著改善系统的稳态性能，简单容易实现，显著提高了DPC的工程实用价值。针对传统模型预测功率控制存在的稳态性能差、开关频率不固定和计算量等缺点，首先提出了基于有功无功误差最小的二矢量MPC，该方法可以显著改善稳态性能而不影响动态效果，具有较大的实用价值；其次针对PWM整流器采用MPC进行优选矢量时存在的计算量大的问题，以复功率共轭的负值为控制变量，在同步旋转坐标系上对矢量选择进行分析和推导，得出一种只需一次预测和比较即可得到最佳电压矢量的低复杂度MPC；进一步将该方法用于二矢量MPC，得到了矢量计算时间的简化表达式，大大简化了算法的复杂度。针对电网不平衡下的PWM整流器控制，基于传统瞬时功率理论和扩展瞬时功率理论提出了一系列创新思想并和DPC、MPC等先进控制方法相结合，无需旋转坐标变换和正负序分解，为非理想电网下PWM整流器控制提供了新型解决方案。