基于EEMD样本熵的三电平逆变器故障诊断方法的研究

发布时间：2018-03-28 15:36

本文选题：三电平逆变器　切入点：EEMD　出处：《安徽理工大学》2017年硕士论文

【摘要】：当今社会电子信息技术的迅速发展,自动化设备故障的发生率也逐渐增多。在高压、大功率的场所中,三电平逆变器作为一个重要的元器件被广泛应用,但是其结构相对复杂,可靠性较低,因而我们不得不重视其故障问题。目前三电平逆变器的故障处理仍不够准确高效。本课题针对目前所用的三电平逆变器,对其故障进行分析,提出了 EEMD样本熵算法对其故障情况进行研究。本文以三电平逆变器的故障诊断为研究课题,研究对象为二极管钳位式三电平逆变器。首先,分析了三电平逆变器功率器件故障的各种情况,从其拓扑结构以及工作原理,对功率开关管的开路故障进行分类,通过MATLAB平台建立故障模型,得到各种故障状态下各相桥臂中点电位电压波形。其次我们用EEMD(集合经验模态分解)算法将故障电压信号波形分解成若干IMF(本征模态函数),选取IMF分量进行样本熵计算,从而得到故障特征向量;之后通过得到的故障特征向量来研究最小二乘支持向量机算法对故障的诊断,其中采用改进的量子粒子群算法对其核函数σ及惩罚参数C进行优化,得到最终的诊断结果,验证了该算法的有效性。最后进行了基于ARM的故障诊断系统总体方案的设计。在硬件上,所设计系统采用S3C2440微处理器芯片作为主控制器,设计了外围硬件电路。在软件上详细描述了构建嵌入式软件平台的各个步骤。证实了本课题研究的算法的可行性。本课题主要针对三电平逆变器的故障诊断进行了研究。提出了一种EEMD样本熵算法,并得到了最后的结果。通过最终结果验证了本算法的准确性高效性,并在最后进行了系统设计。
[Abstract]:With the rapid development of electronic and information technology, the fault rate of automation equipment is increasing gradually. In high voltage and high power places, three-level inverter is widely used as an important component, but its structure is relatively complex. Because of the low reliability, we have to pay attention to the fault problem. At present, the fault treatment of the three-level inverter is still not accurate and efficient. In this paper, the fault of the three-level inverter used at present is analyzed. In this paper, the fault diagnosis of three-level inverter is studied, and the research object is diode clamped three-level inverter. All kinds of faults of three-level inverter power device are analyzed. According to its topology and working principle, the open circuit fault of power switch is classified, and the fault model is established by MATLAB platform. The voltage waveforms of the midpoint potential of each phase bridge arm in various fault states are obtained. Secondly, the waveform of fault voltage signal is decomposed into several IMFs (intrinsic mode function) by EEMD (set empirical Mode decomposition) algorithm, and the IMF component is selected to calculate the sample entropy. The fault eigenvector is obtained, and then the fault diagnosis is studied by the obtained fault eigenvector, in which the improved quantum particle swarm optimization algorithm is used to optimize the kernel function 蟽 and the penalty parameter C. Finally, the overall scheme of fault diagnosis system based on ARM is designed. In hardware, the S3C2440 microprocessor chip is used as the main controller. The peripheral hardware circuit is designed, and each step of constructing embedded software platform is described in detail in software. The feasibility of the algorithm is confirmed. This subject is mainly aimed at fault diagnosis of three-level inverter. A EEMD sample entropy algorithm is proposed. The accuracy and efficiency of the algorithm are verified by the final results, and the system design is carried out at the end.
【学位授予单位】：安徽理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM464;TP18

【相似文献】