基于优化无迹卡尔曼滤波的电网动态谐波检测

发布时间：2018-11-26 17:46

【摘要】：随着电力电子技术的飞速发展,大容量与非线性电子元件在电力系统中的广泛应用会引起电网电压和电流波形的畸变,由此带来的电能质量问题越来越突出,引起了人们的广泛的关注。电网谐波不仅降低了电力设备的利用效率,而且影响用电设备的正常工作,特别是引发起局部电路谐振,使电压升高、谐波放大,危害用户的用电安全。然而,越来越多的敏感负荷,如可编程控制器、计算机和精密仪器等,却对电能质量提出了更高的要求。因此,有必要准确地检测并给出电网谐波参数,从而准确进行电网谐波评估和电网谐波治理。对电网谐波信号进行及时、准确的检测分析,减少由谐波导致的继电保护和自动装置的误动,从而提高电力设备的效率,降低用电成本。本文首先分析了4种常用的电能质量分析方法:有效值法、傅立叶变换法、小波变换法和自适应的最小二乘法,并分别对以上算法进行仿真分析。然后,阐述和分析了卡尔曼滤波、无迹卡尔曼滤波基本原理并分别进行了算例仿真。无迹卡尔曼滤波算法将状态噪声协方差和观测噪声协方差视为常量,不能准确反映实时变化的噪声环境,估计效果差。本文提出利用基于种群分类与动态学习因子的改进粒子群优化算法,对无迹卡尔曼滤波的状态噪声协方差和观测噪声协方差进行优化,结合无迹卡尔曼滤波对电网动态谐波进行估计。给出了基于粒子群优化的无迹卡尔曼滤波(particle swarm optimized unscented Kalman filter,PSOUKF)算法流程,运用MATLAB进行编程,对电网动态谐波估计进行仿真分析,并将本文所提算法与卡尔曼滤波算法、无迹卡尔曼滤波算法进行比较。仿真结果表明,本文所提方法比传统分析方法更有效,在没有增加计算复杂度的情况下,能够提高动态谐波估计精度。
[Abstract]:With the rapid development of power electronics technology, the wide application of large capacity and nonlinear electronic components in power system will lead to the distortion of voltage and current waveforms of power network, and the problems of power quality are becoming more and more prominent. Has aroused the widespread concern of people. Harmonics not only reduce the utilization efficiency of power equipment, but also affect the normal operation of electric equipment, especially the local circuit resonance, which makes the voltage rise, harmonic amplifies, and endangers the safety of users. However, more and more sensitive loads, such as programmable controllers, computers and precision instruments, require higher power quality. Therefore, it is necessary to accurately detect and give the harmonic parameters of the power network, so as to accurately evaluate and treat the harmonic of the power network. In order to improve the efficiency of power equipment and reduce the cost of electricity consumption, the harmonic signals are detected and analyzed in time and accurately to reduce the relay protection caused by harmonics and the misoperation of automatic devices. In this paper, four commonly used power quality analysis methods are analyzed firstly: effective value method, Fourier transform method, wavelet transform method and adaptive least square method, and the above algorithms are simulated and analyzed respectively. Then, the basic principle of Kalman filter and unscented Kalman filter are described and analyzed. The unscented Kalman filter takes the state noise covariance and the observation noise covariance as constants and can not accurately reflect the real time changing noise environment. In this paper, an improved particle swarm optimization algorithm based on population classification and dynamic learning factor is proposed to optimize the state noise covariance and observation noise covariance of unscented Kalman filter. The unscented Kalman filter is used to estimate the dynamic harmonics of the power system. The flow chart of unscented Kalman filter (particle swarm optimized unscented Kalman filter,PSOUKF) algorithm based on particle swarm optimization is presented. The dynamic harmonic estimation of power network is simulated and analyzed by using MATLAB, and the algorithm proposed in this paper and Kalman filter algorithm are presented. The unscented Kalman filtering algorithm is compared. Simulation results show that the proposed method is more effective than the traditional analysis method and can improve the accuracy of dynamic harmonic estimation without increasing computational complexity.
【学位授予单位】：深圳大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN713;TM935

【参考文献】