多相永磁同步电机具有高功率密度、高可靠性、转矩脉动小等优势，被广泛应用于汽车、船舶推进、航空航天等工业领域。其中，相移30°的双三相永磁同步电机与传统三相电机由于结构相似且拥有多相电机的优点，因而得到广泛关注。随着微处理器的快速发展，模型预测控制作为一种先进的控制算法，广泛应用于功率变换器和电机驱动等领域。本文以相移30°的表贴式双三相永磁同步电机为研究对象，对其矢量控制策略、模型预测电流控制策略以及断相故障下的模型预测控制容错策略等问题进行深入研究，并对上述控制策略进行了仿真分析与实验验证。

本文首先分析了双三相永磁同步电机在自然坐标系下的数学模型，基于矢量空间解耦理论推导了坐标变换矩阵，建立矢量空间解耦坐标系下的数学模型，接着分析了六相电压源逆变器的电压矢量分布情况，在此基础上分析了基于矢量空间解耦坐标变换的矢量控制策略。

其次，针对模型预测电流控制策略的计算时间长、电流谐波较大的问题，提出一种无差拍优化的模型预测控制策略。采用无差拍预测与模型预测相结合的方法，在考虑两者一拍延迟补偿的前提下，基于伏秒平衡原则，选取基波子平面中的大矢量和中矢量构建合成虚拟电压矢量作为矢量控制集，以简化预测模型和减少电流谐波。采用无差拍方法对期望电压位置进行估算，通过定位扇区将矢量个数由13个减少为3个，降低了系统计算负担。仿真结果表明，与有限集模型预测算法相比，无差拍优化模型预测算法能够达到更优的动稳态性能，同时相电流的谐波含量减少，也降低了计算量，验证了所提算法的可行性。

然后，针对双三相永磁同步电机的单相断路故障，将提出的无差拍优化的模型预测电流控制策略拓展到容错控制中，采用模型预测算法跟踪容错参考电流，并分析了断相后的基本电压矢量分布情况，提出一种优化的虚拟矢量合成方法，分析了基于简化控制集的模型预测控制容错策略。

最后，搭建了基于TMS320F28335的双三相永磁同步电机驱动系统实验平台，基于理论分析，编写矢量控制与改进前后的模型预测电流控制的程序，并进行实验验证与对比分析。实验结果与仿真结果一致，验证了上述控制策略的可行性。

Polyphase permanent magnet synchronous motors have the advantages of high power density, high reliability, and small torque ripple, and are widely used in automotive, marine propulsion, aerospace and other industrial fields. Among them, the two-phase permanent magnet synchronous motor with a phase shift of 30° and the traditional three-phase motor have received extensive attention because of their similar structure and the advantages of multi-phase motors. With the rapid development of microprocessors, model predictive control, as an advanced control algorithm, is widely used in the fields of power converters and motor drives. In this paper, the surface-mounted dual-phase permanent magnet synchronous motor with a phase shift of 30° is taken as the research object, and its vector control strategy, model predictive current control strategy and model predictive control fault-tolerant strategy under phase failure are studied in depth. Simulation analysis and experimental verification of the above control strategy are carried out.

Firstly, the mathematical model of a dual-phase permanent magnet synchronous motor in the natural coordinate system is examined in this study. The coordinate transformation matrix is derived based on the vector space decoupling theory. And the decoupling vector coordinate system is established. The six-phase voltage is examined. The voltage vector distribution of the source inverter is analyzed. On this basis, the vector control strategy based on the vector space decoupling coordinate transformation is analyzed.

Secondly, in view of the long calculation time and large current harmonics of the model predictive current control strategy, a deadbeat optimization model predictive control strategy is proposed. Using the method of combining deadbeat prediction and model prediction, considering the one-beat delay compensation of the two, based on the principle of volt-second balance, the large vector and the middle vector in the fundamental wave sub-plane are selected to construct a synthetic virtual voltage vector as the vector control set. To simplify prediction models and reduce current harmonics, the deadbeat method is used to estimate the expected voltage position. And the number of vectors is reduced from 13 to 3 through the positioning sector, which reduces the computational burden of the system. The simulation results show that, compared with the finite set model prediction algorithm, the deadbeat optimization model prediction algorithm can achieve better dynamic and steady state performance. Additionally, the harmonic content of the phase current is lowered and the amount of computation required to verify the viability of the proposed technique is also reduced.

Then, the suggested deadbeat optimization model predictive current control approach is extended to fault-tolerant control, aimed at the single-phase open circuit failure of the dual-phase permanent magnet synchronous motor. The model prediction algorithm is used to track the fault-tolerant reference current. The analysis of the fault-tolerant reference current after phase outage is analyzed. Based on the distribution of basic voltage vectors, an optimized virtual vector synthesis method is proposed. The fault-tolerant strategy of model predictive control based on simplified control set is analyzed.

Finally, an experimental platform of dual-phase permanent magnet synchronous motor drive system based on TMS320F28335 is built. Based on theoretical analysis, the program of vector control and model predictive current control before and after improvement is written. And experimental verification and comparative analysis are carried out. The experimental results are consistent with the simulation results, which verify the feasibility of the above control strategy.