【作者】 张艳杰；

【导师】 徐丙垠；

【作者基本信息】 山东大学 ， 电力系统及其自动化， 2011， 博士

【摘要】 在风力发电系统研发过程中,需要使用风力机模拟器驱动发电机,以测试发电机和并网装置在不同风速条件下的响应性能。研发精确、经济的风力机特性模拟技术,对于推动风力发电机技术的开发和应用具有十分重要的意义。实验室环境下的风力机模拟器都是由伺服电动机制成的,其对电动机的要求是：具备宽广的调速范围、良好的动态转速和转矩特性、较高的运行效率和可靠性,并且易于控制。目前,风力模拟器采用的电动机类型有直流电动机、异步电动机或者永磁同步电动机。直流电动机带有电刷和滑环,维护量大,可靠性较差,不适合应用于大功率风力机的模拟。异步电动机矢量控制复杂,控制精度受变化的转子参数影响较大,并且低速运行时电机出力小,容易振荡和不稳定。而永磁同步电动机的制造成本则比较高,且存在退磁问题。开关磁阻电动机(Switched Reluctance motor,简称SRM)是近年来开发成功的新型电动机,其优点是：调速范围宽,运行效率高；电机转子结构简单坚固,功率变换器无直通隐患,可靠性高；低速时出力大,有良好的转矩/转动惯量比；控制方式灵活,通过合理的控制,可以实现与直流电动机类似的转矩-转速特性。这些优势使得SRM很适合用于风力机模拟器,但目前还没有SRM用于风力机模拟器的研究报道。本文研究使用SRM的风力机特性模拟技术。对转速调节模拟方法及其模拟误差模型、基于模糊PI调节器的SRM转速控制法、SRM的直接转矩控制方法进行了较为深入、全面的探讨,并开发了实用的使用SRM的风力机模拟器。主要内容有：1.对目前国内外的风力机特性模拟的研究现状进行了分析总结。探讨了SRM应用于风力机特性模拟的可能性,并提出了SRM相对于其它风力机模拟系统的优势。2.提出一种在软件模拟器中数值求解风力机系统微分方程,控制伺服电动机实时调速跟踪的风力机模拟技术。在时域上建立转速调节模拟风力机特性的误差模型,分析模拟步长和转速对风力机特性模拟误差的影响,提出在风力机软件模拟器中以四阶龙格库塔方法实时求解系统微分方程,以降低模拟误差。3.设计了SRM模糊PI调节器,克服SRM双凸极结构和饱和非线性带来的建模困难,进行转速的动态平滑调节。以开关磁阻电动机准线性模型为基础,将SRD(Switched Reluctance motor Drive,开关磁阻电动机调速系统)整定成典型Ⅱ型系统以确定模拟步长。转速偏差大时模糊调节器工作,转速偏差小时PI调节器作用,以消除稳态偏差。进行了风力机模拟系统单步长转速模拟的数字仿真,结果表明模糊PI调节器相比常规PID调节器具备更好的转速跟踪性能。4.研究了一种基于SRM直接转矩控制(Direct Torque control, DTC)的风力机转矩模拟技术。依据风力机和模拟系统的机械方程,推导出转矩调节模拟风力机的转动惯量补偿方法,以实现风力机动态模拟。实时软件模拟器中执行风力机模型,并以转矩指令控制SRM实时跟踪输出参考转矩,SRM采用直接转矩控制策略调节转矩。5.开发了一种转矩磁链双滞环控制的开关磁阻电动机DTC算法。基于SRM近似转矩模型和试验测定的磁化及转矩特性数据,采用转矩磁链双滞环控制策略,直接控制SRM的转矩,实现使用SRM的风力机特性转矩调节模拟。SRM双滞环控制借鉴异步电机的直接转矩控制策略,对SRM磁链矢量实时空间定位,根据参考磁链和转矩幅值,选择控制电压矢量。采用双位式砰砰控制,原理简单,鲁棒性强。Simulink仿真结果验证了转矩磁链双滞环控制的SRM转矩跟踪的有效性,相比SRM的电流斩波控制,SRM直接转矩控制显著减小了转矩脉动。6.分别建立了风力机与风力机模拟系统小信号线性化模型,对风力机特性的转速调节模拟方案与转矩调节模拟方案在频域上进行研究分析,建立模拟误差的频域特性模型,结合输入风速的Van der Hoven模型功率谱特性,比较两种模拟方案的频域动态误差特性。7.建立了开关磁阻电动机驱动直流发电机的风电系统模拟试验平台,设计了一套基于10kW开关磁阻电动机的风力机模拟器。转速调节模拟试验中,上位机软件根据风速输入和传感器的负载转矩值进行风力机特性模拟计算,通过串口向电机控制器传送转速指令,由模糊PI控制器进行实时转速调节。在风力机特性的转矩模拟试验中,测量了开关磁阻电动机的磁链-电流-位置角特性数据,并通过三次样条函数插值,对电流i积分和对位置角9的微分运算,获取SRM的转矩-电流-位置角特性表格,用于开关磁阻电动机DTC控制。以14kW直流发电机为风力机模拟器负载,采用直接转速控制策略,调节直流发电机励磁,对风力机模拟器进行最大风能跟踪测试。转速曲线和负载转矩数据分析表明,所开发的风力机模拟器实现了对风力机转矩-转速特性和转动惯量的模拟。本文对基于开关磁阻电动机的风力机特性模拟技术进行深入研究。分别采用基于SRD模糊PI控制的转速调节方案和基于SRM直接转矩控制的转矩调节方案对风力机特性进行模拟。基于开关磁阻电动机的风力机模拟器结构可靠稳定,调速范围宽,低速动态性能好,为实验室环境下的风力发电机的研发测试提供了一种有效的模拟手段。更多还原

【Abstract】 In the research and development of the wind generation system under the laboratory condition, it is necessary and important to simulate the wind turbine. As the wind mill equipment is expensive and inconvenient, a motor drive is usually used to emulate the wind turbine characteristics. It drives the generator in the wind generation experiments, in order to develop maxmimum power tracking algorithm and test the grid integration equipment performance. For this reason, the wind turbine simulator (WTS) is very important in wind generation studies.The present wind turbine simulators are mostly based on Direct-Current(DC) motor、Asynchronous motor or Permanent Magnet Synchronous Motor(PMSM). The DC motor is not reliable enough for the maintenance of its brushes and slip rings. It is also unsuitable for the simulation of large capacity wind turbines. Asynchronous motor is complex to implement good torque characteristics and the vector control is influenced by the time-varying rotor coefficients. In addition, its low speed output is small, thus results in the vibrations and unstability. The PMSM is costly in the machine material and is also not suitable in simulaion of the large capacity wind turbines. As the development of the wind generation technology, the servo motor used in wind turbine simulator is ought to be durable、reliable and modulated well in the a wide speed range.The switched reluctance motor (SRM) is solid and reliable due to the simple structure of the rotor: brushless and windingless, thus overcomes the drawbacks of DC motor. Its converter is more reliable than the asynchronous motor. Its torque output is large in low speed scope, potentially better than the asynchronous motor in an overall speed range. Also the motor is much cheeper than the PMSM. In addition, it has several control items, adaptive to flexible control and could acquire the mechenical performance as good as the DC motor through approriate control. All these priorities make the SRM a good choice in wind turbine simulation. However, there have been few researches on the wind turbine simulator adopting switched reluctance motor. At present, most of the wind turbine simulators are implemented by the way of torque control, which is hard and complex to realize. It is necessary to make a thorough study in the WTS and explore some new schemes to simplify the WTS structure in the wind generation studies.This paper focuses mainly on the control of the switched reluctance motor to simulate the wind turbine, and has done all-around study, from simulation to experiment, from theory to field practice. The detailed research work is as follows:1. This paper summerized and analyzed the present technology of wind turbine simulation, explored the possibility of the wind turbine simulation based on the switched reluctance motor. The advantages of SRM based WTS is also presented.2. The paper proposes an WTS:The system differential equation is solved in every time stamp to acquire the reference rotational speed the real turbine will arrive at the next time point. The speed commands are sent instantly to SRM controller, which regulate the SRM to transit to the reference speeds. A simulation error model is established to measure the WTS adopting speed regulation scheme. The wind turbine simulation error in function of the simulation step and feedback rotational speed is studied through MATLAB numerical analysis. The 4th Runge-Kutta method is adopted in the solving equation work to reduce the simulation error.3. A fuzzy logic algorithm combined with PI controller is designed to implement the speed regulation scheme based on switched reluctance motor. Because the SRM is highly nonlinear due to the double salient structure and saturated magnetic circuit, as well as completely dependent on switches in normal work, the accurate model is hard to obtain. The WTS in the speed regulation scheme can be realized in the form of fuzzy logic of the chopped current control of SRM. A PI control scheme is added to the fuzzy algorithm in order to eliminate the steady error. The SRM tracking system is set to typical II system to determine the simulation step and the simulation error is measured by accounting the numerical results. Through digital simulation of the WTS in a single step in MATLAB/Simulink environment, the fuzzy PI controller is proved to have a better speed tracking performance than the normal PID controller.4. A wind turbine simulator based on the DTC (Direct torque control) of switched reluctance motor is proposed. In order to emulate the wind turbine dynamically, a turbine inertial compensation method is deduced according to the system mechanical equations and applied to the proposed WTS. The wind turbine model is implemented in the real-time software simulator and the reference torque is acquired and sent to motor controller, which make the SRM track the reference torque in the DTC manner.5. Based on the approximate torque model, the magnetization and torque characteristic data, a dual hysteris loop control method is conducted to implement the DTC of SRM, in purpose of simulating the wind turbine in a torque control scheme. Both the torque and flux amplitudes are controlled in respective target hysteris loops, thus the direct torque control of SRM is carried out. Nowadays the high property control of SRM torque mainly incudes:torque sharing function(TSF)、sliding mode variable structure control、iterative learning and neural network control.The methods are always complex, in need of large amount of off-line computation. The proposed direct torque control algrithm is based on control of dual hysteris loops of flux and torque, which borrows the DTC scheme of the asynchronous motor, and has a double on-off regulator. It locates the flux space vector, calculates the amplitude of the flux and motor torque, then choose the voltage space vector to decrease or increase the amplitude. Consequently, the flux and torque amplitudes are constrained in hysteris bands. The control structure is simple and robust and effective in the torque ripple reduction of SRM. The turbine model combined with the DTC tracking system of SRM is simulated in the MATLAB/Simulink environment. In constant wind, the simulation results show that the dual-hysterisis torque tracking system has good accordance with the real wind turbine characteristics and indicates a much less torque pulsation than the conventional chopped current hysterisis control.6. The wind turbine linearized model and the WTS linearized models both in speed control scheme and torque control scheme is deduced to compare the speed control scheme with the torque control scheme in frequency characterics. A simulation error model is established in frequency domain to evaluate simulation error for the two schemes. With respect to the power spectral characteristics of Van der Hoven modle for wind speed, the simulation error for both schemes is analized and compared. In conclusion, the torque control scheme of wind turbine simulator indicates a better tracking performance in high frequency simulation occation than the speed control scheme, while in the low-medium frequency scope the two schemes have the close simulation error.7. A test rig for the wind turbine simulator based on SRM is built. It is composed of a 10kW SRD (Switched Reluctance motor Drive) and a 14kW DC generator acting as the the load. Firstly, the fuzzy-PI controller of SRD is designed. In the experiments of WTS adopting the speed control scheme, the upper machine software implements the wind turbine model and sends the reference speeds to the SRD controller through RS232 port. The fuzzy-PI controller regulates the SRM according to the track the received speeds. The experimental results suggest that the fuzzy-PI controller of SRD is capable of speed tracking in WTS adopting the speed control scheme. In the experiments for the WTS based on DTC of SRM, The flux-current-position angle characteristics are measured and the flux-current-position angle characteristics are deduced by integrating the current i and differentiating the position angelθ。The charateristics are stored as data forms in the program and used to calculate the current fluxes and torques. The control program is made up of two parts:the wind turbine model implementation and the DTC algrithm of SRM torque track. The experimental results indicate that the WTS can perform the wind turbine characteristic under constant wind velocity with variable load. In the MPPT (Maximum Power Tracking point) test, the WTS characteristic also has a good identity with the simulated wind turbine.The thesis makes intensive study on the wind turbine simulator based on switched reluctance motor. The speed control scheme in WTS adopting the fuzzy PI control of SRM is put forward, as well as the direct torque control scheme of SRM in torque control scheme of WTS. The speed control scheme is simple, reliable and has a good stability. The torque control scheme of WTS is comparatively complex while the simulation error in high frequency domain is smaller than the speed control scheme. In conclusion, the torque control scheme of WTS is an ideal method. The wind turbine simulator based on switched reluctance motor indicates a reliable merit, high range speed regulation and good performance in low speed range. It is an effective servo-motor for the development of wind generation systems. The proposed WTS based on switched reluctance motor has provided a practical instrument for the wind generation study and it exhibits a prospect in large quantity and direct drive wind generation developments. Due to the intrinsic torque pulsation and heavy nonlinearity of switched reluctance motor, it is hard to modulate the torque and speed accurately. The torque pulsation minimization study is ought to be conducted to ensure and improve the performance of the wind turbine simulator in future work.更多还原