电动汽车IPMSM驱动系统集成故障在线诊断与容错控制研究

The research on the interior permanent magnet synchronous motor (IPMSM) drive system integrated online fault diagnosis and fault-tolerant control is an urgent needed technology to achieve safe and reliable operating of electric vehicle (EV), and seeking for an online fault diagnosis method and fault-tolerant control technology is the important way to solve the problem. Based on the modeling of IPMSM drive system in healthy state and fault state allowing for the measurement noise and modeling uncertainty, the research is committed to achieve the online diagnosis of balanced and partial demagnetization fault by the unscented Kalman particle filter and Hilbert-Huang Transform based on adaptive fundamental wave extraction, and achieve the adaptive fault-tolerant control of demagnetization fault by solving the adaptive characteristics inter-harmonic suppression technology of partial demagnetization and the maximum torque per ampere (MTPA) and flux-weakening control technology of different demagnetization degree of failure for vector controlled IPMSM system. Moreover, the research is committed to achieve the unified online diagnosis of key sensor faults based on the full-decoupled parity space method and the generalized likelihood estimation, and achieve the adaptive fault-tolerant control by reveal the essential relationship between the different position sensor faults and the nonlinear characteristics of stator current. And then, the electric drive system of EV is build up integrated IPMSM drive control with online fault diagnosis and fault-tolerant control, the abovementioned researched aims to open up the new ways of implementing the safe and reliable operation of EV and make the new contribution for improving the technical level of EV integrated online fault diagnosis and fault-tolerant control.

IPMSM驱动系统集成故障在线诊断与容错控制研究是实现电动汽车安全可靠运行的迫切科技需求，寻求故障在线诊断方法及容错控制技术是解决这一问题的重要途径。本项目基于考虑测量噪声和建模不确定性的IPMSM驱动系统健康及故障状态建模，研究无迹卡尔曼粒子滤波和自适应基波提取的希尔伯特-黄变换相结合实现永磁体均匀和非均匀退磁故障的在线诊断；攻克非均匀退磁特征间谐波自适应抑制技术和不同退磁程度下IPMSM矢量控制系统最大转矩电流比控制和弱磁控制技术实现退磁故障的自适应容错控制。研究完全解耦的等价空间法和广义似然估计相结合实现关键传感器故障统一在线诊断，揭示不同位置传感器故障与定子电流表征出的非线性特征间的本质联系实现其自适应容错控制。架构集成IPMSM驱动控制与故障在线诊断与容错控制为一体的电动汽车电驱动系统，开辟电动汽车安全可靠运行的新途径，为提升电动汽车集成故障在线诊断与容错控制技术水平做出新贡献。

给出了横向磁通永磁电机（TFPM）兼顾漏磁与齿槽转矩抑制的设计流程，攻克了多目标优化算法和目标函数评价策略选择、优化设计变量和目标函数确定、TFPM多目标优化设计流程实现等多项关键技术，提出了TFPM设计及多目标优化完整的解决方案。.创新性设计出双输入双输出无模型自适应控制器(MFAC)，根据实时观测转矩且利用双梯度下降法生成定子电流指令，基于电流闭环控制实现了PMSM驱动系统全速度范围内的精确转矩控制。创新性设计出容错PI调节器饱和故障的PMSM弱磁控制器，实现了PMSM驱动系统的弱磁扩速安全运行。.将无模型控制与无差拍预测控制相结合，基于PMSM驱动系统的输入输出建立系统的超局部模型，再设计出无模型无差拍电流预测控制器，一体化解决了具有参数不确定性与逆变器非线性的PMSM驱动系统高性能实时控制难题。.提出了基于代数法的PMSM多参数辨识方法，解决了电机参数变化对永磁体磁链辨识精度的影响及非线性滤波方法在多参数同时辨识中所存在的辨识模型欠秩问题，实现了永磁体磁链的准确辨识。提出了自适应基波提取和经验模态分解（EMD）重构滤波相结合，基于分形维数的PMSM永磁体局部退磁故障诊断方法，实现了PMSM永磁体局部退磁故障的准确诊断及故障程度的准确刻画。.提出了集成逆变器死区补偿、电流传感器故障诊断与容错控制的PMSM驱动系统安全稳定运行的解决方案，实现了电流传感器故障的在线准确诊断与容错控制。所提出的位置传感器故障综合解决方案，不仅能实现位置传感器故障的在线准确诊断与容错控制，而且巧妙地回避了逆变器死区效应对位置传感器故障在线诊断与容错控制的影响。.该项目的研究成果，对于实现电动汽车PMSM驱动系统在宽调速范围内的安全可靠运行兼具重要的理论研究价值和工程应用价值，具有广阔的市场应用前景。该项目培养出培养硕士14名，博士2名，在国内外期刊、国际会议上发表论文28篇，其中SCI期刊论文3篇，EI期刊论文8篇，获批国家发明专利8项，建设完成PMSM驱动系统实验平台。