可视化智能配电网的预警研究

发布时间：2018-05-02 18:23

本文选题：可视化 + 关联规则挖掘　；参考：《东北大学》2014年硕士论文

【摘要】：随着我国经济的蓬勃发展,用户对电能质量的要求也越来越高,配电网的安全运行变得非常重要。配电网的可视化预警可以直观、快速的显示系统中的异常情况,将配电网的实时动态反映给电网运行和管理人员,防止异常状态的恶化,保障智能配电网安全、稳定、可靠的运行。本文首先根据配电网SVG文件,对配电网网架结构进行解析与重构,之后以构建了一种配电网预警评估的方法。本文将关联规则挖掘方法引入配电网的关联度评估中,根据配电网预警的需要对AprioriTid算法进行优化,应用优化后的算法得到元件的关联度。考虑到灾害气候对配电网元件的巨大影响,本文将灾害气候数据与配电网运行相关数据结合,利用相同方法进行关联规则挖掘,进而计算故障概率。合成上述两种参数和可靠性评估参数,得到元件的预警评估结果。在预警评估的基础上,使用基于信息熵和关联规则挖掘的免疫算法和量子免疫算法两种免疫算法得出预警结果。本文给出了免疫算法应用在预警中的具体流程,并对算法进行了分析。文章的最后以实际的配电网为例子,使用软件验证了本文提出的预警方法。经验证,本文提出的预警方法得出的预警结果与仿真结果大致一致,预警效果较好。本文所提出的配电网网架结构解析与重构方法和配电网预警方法已应用于东北大学与国家电网辽宁省电力有限公司合作项目“提高配电网故障处理能力关键技术研究与开发”的软件中。
[Abstract]:With the rapid development of economy in China, the requirement of power quality is becoming higher and higher, and the safe operation of distribution network becomes very important. The visual early warning of distribution network can directly and quickly display the abnormal situation in the system, and reflect the real-time dynamic situation of the distribution network to the network operation and management personnel, prevent the deterioration of the abnormal state, and ensure the security and stability of the intelligent distribution network. Run reliably. In this paper, the distribution network structure is first analyzed and reconfigured according to the distribution network SVG file, and then a method of early warning evaluation of distribution network is constructed. In this paper, the association rules mining method is introduced into the evaluation of the distribution network correlation degree, and the AprioriTid algorithm is optimized according to the need of the distribution network early warning, and the correlation degree of the components is obtained by using the optimized algorithm. Considering the great influence of disaster climate on distribution network components, this paper combines disaster climate data with relevant data of distribution network operation, uses the same method to mine association rules, and then calculates the failure probability. The two parameters mentioned above and the reliability evaluation parameters are synthesized, and the early warning evaluation results of the components are obtained. On the basis of early warning evaluation, two immune algorithms based on information entropy and association rule mining and quantum immune algorithm are used to get the early warning results. In this paper, the flow chart of the application of immune algorithm in early warning is given, and the algorithm is analyzed. At the end of this paper, we use the software to verify the proposed early warning method, taking the actual distribution network as an example. It is verified that the early warning results obtained by the proposed method are approximately consistent with the simulation results, and the early warning effect is good. The analysis and reconfiguration method of distribution network structure and the early warning method of distribution network proposed in this paper have been applied to the cooperative project of Northeast University and State Grid Liaoning Electric Power Co., Ltd. "the key technology of improving the fault handling capacity of distribution network" Research and development of the software.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM727;TP311.13

【参考文献】