随着电网技术的发展与“双碳”发展目标等绿色环保理念的加深，大量分布式电源（Distributed Generation, DG）接入配电网，导致配电网网络结构与故障特征更为复杂，潮流分布双向多变，且考虑到恶劣环境、天气等因素致使馈线终端单元（Feeder Terminal Unit, FTU）上传的故障信息容易出现错漏等畸变情况，使得传统配电网故障定位方法失效。因此，本文在传统故障定位方法的基础上，对适用于有源配电网的故障区段定位方法展开研究。
       首先，本文介绍了 DG 的概念以及并网原则，分析了 DG 接入配电网的不同位置对配电网故障区段定位的影响；针对 DG 并网导致传统配电网编码方式、开关函数和评价函数不再适用的问题，提出了新的有源配电网编码规则、开关函数和评价函数模型，并用具体算例验证了模型可行性，为下文研究有源配电网故障区段定位方法提供了理论基础与依据。
       其次，选择多元宇宙优化算法（Multi-verses Optimization, MVO）用于解决有源配电网故障区段定位问题，针对传统 MVO 收敛速度较慢、易陷入局部最优解等缺点，本文设计了相应的优化策略，对算法重要参数的变化规律、种群初始值等进行改进，引入压缩因子和混沌映射策略增强算法寻优搜索能力，以此提升算法的整体性能。采用典型测试函数进行算法性能检测，结果表明改进后的多元宇宙优化算法（Improved Multi-verses Optimization, IMVO）相较于其它智能优化算法在寻优性能与迭代速度方面具有突出优势。本文还设计了基于 IMVO 的有源配电网故障区段定位模型,并通过故障区段定位仿真实验验证了所提方法在故障区段定位方面的优势。
       最后，为了解决有源配电网复杂故障情况下故障区段定位准确率和效率较低的问题，本文借鉴配电网分层理论，结合 IMVO 与枚举法（Exhaustion Method, EM），提出了基于IMVO+EM 的有源配电网分层故障区段定位方法，通过将配电网故障区段定位分为区域定位与区段定位两部分，降低求解难度，提高定位准确率与效率。通过仿真实验对方法
有效性进行了验证，仿真定位结果表明，基于 IMVO+EM 的有源配电网分层故障区段定位方法针对有源配电网复杂故障情况具有较高的故障定位准确率与效率，定位性能优越。

      With the development of power grid technology and the deepening of green environmental protection concepts such as the "double carbon" development goal, a large number of Distributed Generation (DG) are connected to the distribution network, resulting in more complex network structure and fault characteristics, and bidirectional power flow distribution. Moreover, the fault information uploaded by the Feeder Terminal Unit (FTU) is prone to distortion such as error and leakage due to factors such as harsh environment and weather, which makes the traditional distribution network fault location method ineffective. Therefore, based on the traditional fault location method, this paper studies the fault section location method suitable for active distribution network.

      Firstly, this paper introduces the concept of DG and the principle of grid connection, and analyzes the influence of different locations of DG access to distribution network on fault location of distribution network. To solve the problem that the traditional distribution network coding method, switching function and evaluation function are no longer applicable due to the grid-connected DG, a new active distribution network coding rule, switching function and evaluation function model is proposed, and the feasibility of the model is verified by a concrete example, which provides a theoretical basis and basis for the following research on fault section location method of active distribution network.

       Secondly, Multi-verses Optimization (MVO) algorithm is selected to solve the fault section location problem of active distribution network. In view of the shortcomings of traditional MVO, such as slow convergence speed and easy to fall into the local optimal solution, this paper designs the corresponding optimization strategy. The change rule of important parameters and initial population value of the algorithm are improved, and compression factor and chaos mapping strategy are introduced to enhance the optimization and search ability of the algorithm, so as to improve the overall performance of the algorithm. The performance of the algorithm is tested by using typical test functions. The results show that the Improved Multi-verses Optimization (IMVO) algorithm has outstanding advantages in optimization performance and iteration speed compared with other intelligent optimization algorithms. A fault section location model based on IMVO is also designed in this paper, and the advantages of the proposed method in fault section location are verified by simulation experiments.

      Finally, in order to solve the problem of low accuracy and efficiency of fault section location under complex faults of active distribution network, this paper draws on the hierarchical theory of distribution network and combines IMVO and enumeration Method (EM) to propose a hierarchical fault section location method of active distribution network based on IMVO+EM. By dividing the fault location of distribution network into two parts: regional location and regional location, the difficulty of solution is reduced, and the accuracy and efficiency of location are improved. The simulation results show that the layered fault section location method based on IMVO+EM has high fault location accuracy and efficiency, and excellent location performance for complex faults in active distribution network.