智能配电网故障定位及保护配合的研究

发布时间：2019-04-25 18:36

【摘要】：21世纪以来,我国电力行业迅猛发展,用户对供电质量及可靠性的要求逐日提高。配电网是电力用户与发输电的重要纽带,若其发生故障将导致配网局部断电,更有甚者致使全网断电。我国配网系统主要采用小电流接地方式,单相接地故障占到总故障的60%甚至更高,而准确快速实现单相接地故障的选线和定位可以有效缩短停电时间,减少人力消耗,具有重要现实意义,也是电力部门与科研工作者努力研究的方向之一。本文在分析了配电网单相接地故障暂稳态特征基础上,研究了配电网小电流单相接地故障的暂态零序电流区段特征,并结合故障点两侧电流突变方向不同及小波变换对奇异信号有较强的识别能力,构造了一种利用小波变换模值极性的故障区段判别方法。应用PSCAD/EMTDC仿真软件对复杂配网进行仿真分析,结果表明该方法能良好实现故障区段的定位。在故障测距方面,因配网小电流故障信息微弱,利用单一故障信息的定位方法易受外界因素影响,故构建了以零序电流基波、暂态小波能量、暂态无功功率为输入特征量的多故障信息小波神经网络测距模型,并采用免疫粒子群算法优化小波神经网络的各参数初始选择,得到各参数的最优初始值,以减少选择随机数带来的是否收敛问题。MATLAB仿真结果表明,优化后的小波神经网络定位方法与BP神经网络、小波神经网络定位方法相比,能有效提高学习效率和测距精度。最后简要分析了基于配网自动化的分布式与集中式控制保护理论。
[Abstract]:Since the 21 ~ (st) century, the electric power industry in our country has developed rapidly, and the demand for the quality and reliability of power supply is increasing day by day. Distribution network is an important link between power users and power generation and transmission. If its fault occurs, it will lead to the local power loss of the distribution network, and even more, it will lead to the power loss of the whole network. Single-phase grounding fault accounts for 60% or more of the total fault in China's distribution network system, and the accurate and rapid line selection and location of single-phase grounding fault can effectively shorten the blackout time and reduce the manpower consumption, and the fault line selection and location of the single-phase grounding fault can effectively shorten the blackout time and reduce manpower consumption. It has important practical significance, and it is also one of the research directions of electric power department and scientific research workers. Based on the analysis of transient characteristics of single-phase earthing fault in distribution network, this paper studies the characteristics of transient zero-sequence current section of small-current single-phase grounding fault in distribution network. Combined with the different direction of current sudden change on both sides of the fault point and the strong ability of wavelet transform to identify singular signals, a fault section discrimination method using modulus polarity of wavelet transform is proposed. The PSCAD/EMTDC simulation software is used to simulate and analyze the complex distribution network. The results show that this method can realize the location of fault section well. In fault location, because the fault information of small current in distribution network is weak, the locating method based on single fault information is easy to be influenced by external factors, so the zero sequence current fundamental and transient wavelet energy are constructed. A multi-fault information wavelet neural network ranging model with transient reactive power as input characteristic is proposed. Immune particle swarm optimization algorithm is used to optimize the initial selection of the parameters of the wavelet neural network, and the optimal initial values of each parameter are obtained. MATLAB simulation results show that compared with BP neural network and wavelet neural network localization method, the optimized wavelet neural network localization method can effectively improve the learning efficiency and ranging accuracy. Finally, the distributed and centralized control and protection theory based on distribution automation is analyzed briefly.
【学位授予单位】：中国石油大学(华东)
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM862

【相似文献】