【作者】 史婷娜；

【导师】 万健如；

【作者基本信息】 天津大学 ， 电力电子与电力传动， 2009， 博士

【摘要】 本文以低速大转矩永磁同步电机作为研究对象,从电机设计与控制技术相结合的角度出发,着重进行了永磁同步电机驱动控制和优化设计两方面的研究。首先基于坐标变换原理,详细分析了永磁同步电机在静止三相坐标系、静止两相以及旋转两相坐标系下的数学模型,深入讨论了永磁同步电机电压空间矢量控制方法,提出了基于矩阵变换器的永磁同步电机空间矢量调制策略。根据矩阵变换器的拓扑结构特点,将交-交直接变换结构矩阵变换器等效为由“虚拟整流器”和“虚拟逆变器”组成的交-直-交结构,深入分析了矩阵变换器的变换关系,推导出矩阵变换器的电压、电流双空间矢量PWM调制函数。为了验证基于矩阵变换器的永磁同步电机空间矢量调制策略的可行性,以及调制方法的正确性,首先使用Matlab软件平台搭建了空间矢量调制的仿真模型,并以低速大转矩永磁同步电机作为被控对象,对矩阵变换器的输入输出特性进行分析。仿真结果表明,矩阵变换器输出电压的幅值和频率独立可调,实现了电机在宽广范围的连续可调、任意负载下的输入功率因数为1,且能实现能量的回馈,克服了传统交-直-交变频器驱动永磁同步电机所存在的输入谐波大、输入功率因数低、直流回路需要大的储能电容等缺点。解决了传统交-交变频器存在的功率因数低、谐波污染严重以及调速范围窄等问题,改善了永磁同步电机调速系统的运行性能。在仿真的基础上,搭建了基于矩阵变换器的永磁同步电机实验系统,该系统以DSP为主控制器进行空间矢量调制,以FPGA为协处理器实现矩阵变换器开关函数的合成和四步换流策略的实施,成功实现了对输出电压和输入电流的同时调制,获得期望的正弦输入电流和输出电压,保证了永磁同步电机优良的起动和调速性能。实验结果表明,该控制器不仅具有永磁同步电机驱动系统调速精确、动态响应快、鲁棒性强等特点,同时还具有输入、输出电流谐波小、无中间滞留滤波电容、能量再生等特点。针对电机优化设计中存在的高度非线性问题,本文提出了一种基于模糊专家系统的多目标粒子群算法,根据专家设计经验,构建模糊专家系统,将专家经验融合到粒子群速度方程中,引导支配子集更快地向最优解方向运动,改善了多目标粒子群算法寻优过程,提高了求解精度和收敛速度。同时,将电机优化问题转化为博弈问题,提出基于博弈论的电机优化设计方法,从纳什均衡解出发,运用策略性让步博弈的思想综合考虑各优化目标,从非劣解集中搜索综合最优解,降低了电机开发难度、缩短了设计周期,为实现电机设计智能化奠定了基础。针对加工、装配过程中产生的误差及实际应用中存在的干扰因素,在电机设计中引入稳健设计的思想,采用信噪比方法分析了设计方案的稳健性,建立了评价和排序非劣解的重要指标。同时基于灵敏度分析思想,揭示了电机各设计参数与优化目标的内在关系,为电机生产和设计提供了理论指导。最后采用基于模糊专家系统和博弈论的多目标粒子群算法对电梯用低速大转矩永磁同步电机进行了优化设计,测试和应用结果表明,优化设计后,电机综合性能指标明显提高。更多还原

【Abstract】 Centreing around the permanent magnet synchronous motor, two parts of researches on the motor optimization design and the drive control are systematically introduced in this dissertation.Based on the theory of axes transformation, the mathematic models of PMSM in three-phase stationary coordinates, two-phase stationary coordinates and two-phase rotating coordinates are analyzed in details. SVPWM method is introduced in the second chapter. The AC-AC transformation control and the double SVPWM scheme of the matrix converter are deduced from the equivalent structure of the AC-DC-AC conversion in the third chapter.In order to validate the feasibility of the SVPWM strategy for permanent magnet synchronous machine based on matrix converter and the correctness of the modulation method, the simulation model of vector control speed regulating system for a lower- speed high-torque PMSM is firstly established based on Matlab. Then the input and output characteristics of matrix converter, the speed, torque and load characteristics of the PMSM are then analysed. The simulation results show that it can realize the separate modulation of the output voltage amplitude and frequency. Besides, the excellent waveforms of input current and output voltage, unit value power factor reveal the matrix converter system has the ability of four guardant operations. It has unexampled merits on controllable input power factor, power regeneration and lack of bulky energy storage capacitors.After simulation, a PMSM control system based on matrix converter is experimented, which uses DSP TMS320F2812 as main controller and FPGA EP1C6 as coprocessor. DSP carries out space vector modulation, and FPGA carries out switch functions synthesis and four step commutations. They together accomplish the modulation of sine wave input current and output voltage. The excellent startup characteristic, high control precision, quick dynamic response and good adaptability are ensured which can be seen from the experiment results. All the data indicate again that the method in this dissertation is effective.On the subject of high nonlinear existing in optimization design, a sort of multi-objective particle swarm optimization algorithm based on fuzzy expert system is proposed in the fourth chapter. The man expert experiences on design are imported into the artificial fuzzy expert system, and then join in the speed equations of the particle swarm. Thus lead the dominant subsets move to the directions of optimal solution. Fuzzy expert system improves the optimization process of multi-objective particle swarm optimization algorithm, and advances the convergent rate and solving precision. The other novel point in this dissertation is that the optimization problem is translated into game problem. The optimized goals are regarded as the game individuals, an optimization design methods of PMSM based on game theory is proposed. Surrounding Pareto optimum under Nash equilibrium, the concession negotiation strategy between the game individuals is applied to the optimized goals in order to get a better global negotiation result. The motor design cycle is simplified by searching the unitary optimal result from non-inferior solution set. For the error due to machining and assembly and the disturbance, the robustness design idea is introduced in motor design. The non-inferior solutions are evaluated and arranging by signal noise ratio. The sensitiveness analysis idea is also introduced in motor design to reveal the interrelation between every design parameter and the optimized goals. At last, a PMSM is optimized by multi-objective particle swarm optimization algorithm based on fuzzy expert system. The result shows that the design method has the merit of high efficiency, and the integrated characteristic of the motor after optimization design is improved.更多还原