摘要
针对传统单传感器直流母线采样系统中相电流重构盲区问题,该文分析传统空间矢量脉冲宽度调制(SVPWM)技术的电流采样方法和重构盲区产生的原因,精确地划分电压空间矢量平面的各个区域。提出一种互补非零矢量脉宽调制(CNSVPWM)方法,通过在中调制区添加互补非零矢量来延长电流采样窗口的时间,满足直流母线电流最小采样时间要求,消除了电流重构盲区,拓展了电机的运行范围。仿真及实验结果表明,采用CNSVPWM方法电流总谐波畸变率(THD)低于6.15%,重构误差低于3%,电机在动态工况下运行正常。
With the development of power electronics and control theory,AC drive systems are widely used in industrial production,transportation,military aerospace,and other fields.In the AC servo control system,the motor phase current is the core parameter to the vector control,and the detection and feedback of the phase current are of great significance in improving the control system’s performance.The traditional vector control system needs at least two current sensors installed on the AC output side through the space voltage pulse width modulation(SVPWM)method to achieve phase current acquisition and real-time feedback.Multiple sensors will inevitably increase the system cost and equipment volume,which is not conducive to the development of control equipment miniaturization.Furthermore,the inconsistencies between multiple sensors will deteriorate the performance of the system.To improve these problems,based on the DC bus single sensor sampling principle and traditional SVPWM method,this paper proposes a complementary non-zero vector pulse width modulation(CNSVPWM)method.CNSVPWM utilizes the DC bus single sensor to carry out the phase current time-sharing acquisition under the action of the complementary non-zero vectors and complete the reconstruction of three-phase currents.Thus,accurate feedback on current information is achieved while reducing the cost and volume of the control system.Firstly,according to the switching states of the three-phase two-level inverter,six effective voltage vectors and two zero vectors are defined,and the corresponding relationship between the DC bus current idc and the phase current can be obtained.Then,the characteristics of the switching devices and the establishment process of the DC bus current are analyzed.The shortest time for the DC bus current to reach a steady state is obtained.That is,the minimum sampling time Tmin for current sampling.According to the minimum sampling time,each sector is divided into the observable area,sector boundary,and low modulation area.The effective voltage vector not meeting the minimum sampling time requirement is defined as a current reconstruction blind area,where the current cannot be completely reconstructed.Aiming at the current reconstruction blind area and based on the volt-second balance principle,the CNSVPWM method utilizes a pair of complementary non-zero vectors to extend the voltage vector action time.For the CNSVPWM method,the duty cycle of effective vectors in each carrier period TS does not change,and the symmetry of the PWM is retained.In the open-loop experiments,the PWM carrier frequency is set to 10 kHz,the sampling frequency is 20 kHz,and the modulation index is M=0.55.MDA805A motor drive analyzer is used to collect and analyze the experimental data.In sector switching,the CNSVPWM waveform is normally sent,and two sample pulses can be accurately triggered.When the reference voltage vector enters the sector boundary area,the three-phase current can be completely reconstructed,and the reconstruction error is 2.72%.To further verify the effectiveness of the proposed method,the experiments are carried out under two dynamic conditions of motor acceleration and deceleration.When the rotational speed changes,the reconstructed phase current can follow the measured phase current in real-time.The FFT results show that compared with the SVPWM method,the current harmonic THD of CNSVPWM is increased by 2.04%,from 4.11%to 6.15%.The following conclusions are drawn through experiments:(1)The CNSVPWM method retains the symmetry of the PWM waveform and does not affect the duty cycle.(2)The CNSVPWM extends the current sampling window and realizes the phase current reconstruction in the middle modulation area.(3)The current THD of CNSVPWM is less than 6.15%,reconstruction error is less than 3%.
作者
申永鹏
武克轩
吴成中
梁伟华
张细政
Shen Yongpeng;Wu Kexuan;Wu Chengzhong;Liang Weihua;Zhang Xizheng(School of Electrical and Information Engineering Zhengzhou University of light industry,Zhengzhou 450002 China;RVC-National Engineering Lab Hunan University,Changsha 410082 China;College of Electrical and Information Engineering Hunan Institute of Engineering,Xiangtan 411104 China)
出处
《电工技术学报》
EI
CSCD
北大核心
2023年第8期2126-2135,共10页
Transactions of China Electrotechnical Society
基金
国家自然科学基金面上项目(62273313,62173134)
河南省科技攻关项目(222102240005)
河南省青年骨干教师培养计划项目(2021GGJS089)
郑州市协同创新专项项目(2021ZDPY0204)
河南省自然科学基金项目(222300420373)资助。