微电网作为新型电力系统的重要组成部分，不仅能够提高电力供应的灵活性与可靠性，还能有效缓解传统配电网调节和消纳分布式电源的压力，推动能源结构的低碳转型。然而，微电网的运行方式灵活，并且具有多分布式电源接入、潮流双向流动及故障电流较小等特征，这导致传统保护方法难以满足微电网的安全运行需求。行波作为一种暂态信号，凭借其特征能快速且准确地识别故障。本文结合故障电压前行波的能量变化率特征和指数系数特征，构造了一种微电网保护方法。主要研究内容如下：

（1）分析了微电网故障电压前行波特征。对暂态行波信号的产生及传播过程进行研究，并得到了系统故障时故障电压前行波能量变化率特征。利用网格法推导微电网区内外故障时故障电压前行波时域表达式，对这两种情况下的故障电压前行波时域表达式进行了分析，得到了故障电压前行波指数系数特征。仿真验证了故障电压前行波能量变化率特征和指数系数特征在并网和孤岛两种运行方式下的普适性。故障电压前行波特征的研究为行波原理运用于保护领域奠定了理论基础。

（2）提出了一种基于故障电压前行波能量变化率偏差的自适应启动判据。根据故障电压前行波能量变化率特征，引入滑动时间窗算法，构造了能量变化率偏差，从而判断故障是否发生。所提判据能有效反映故障电压前行波能量变化率的突变并实现自适应整定阈值，解决了传统启动判据无法正常启动和适应性不强的问题。通过仿真验证了所提自适应启动判据在不同故障位置、不同故障类型和不同运行方式下的可行性。

（3）构造了基于列文伯格-马夸尔特（Levenberg-Marquardt，L-M）法的故障电压前行波指数系数特征的保护动作判据，提出了一种基于故障电压前行波特征的微电网保护方法。通过L-M法对故障电压前行波进行拟合并辨识指数系数，获得指数系数最大值。利用故障电压前行波指数系数特征进行故障识别，构造保护动作判据。再结合基于故障电压前行波能量变化率偏差的启动判据，构造了基于故障电压前行波特征的保护方法，设计了具体保护流程。所提方法不再依赖行波波头信息，能够有效避免因行波波头提取困难而导致的保护拒动或误动问题。仿真结果表明，基于故障电压前行波特征的微电网保护方法在不同故障位置、不同故障类型和不同运行方式的条件下，均能准确识别故障，从而验证了所提方法的可行性和可靠性。

As an important part of the new power system, microgrid can not only improve the flexibility and reliability of power supply, but also effectively alleviate the pressure of traditional distribution networks to regulate and consume distributed power sources, and promote the low-carbon transformation of energy structure. However, the flexible operation mode of microgrids and their characteristics such as multi-distributed power supply access, flow in both directions and small fault current make it difficult for traditional protection methods to meet the safe operation requirements of microgrids. As a transient signal, traveling wave can quickly and accurately identify faults by virtue of its characteristics. In this thesis, a microgrid protection method is proposed by combining the fault voltage forward traveling wave characteristics of the energy rate of change and the exponential coefficient characteristics. The main research contents are as follows:

(1) The microgrid fault voltage forward traveling wave characteristics are analyzed. The generation and propagation process of transient traveling wave signals are investigated to obtain the characteristics of the rate of change of the fault voltage forward traveling wave energy during system faults. The mesh method is used to derive the time-domain expression of the fault voltage forward traveling wave when internal fault and external fault, and the time-domain expression of the fault voltage forward wave in these two cases is analyzed to obtain the fault voltage forward traveling wave exponential coefficient characteristics. The simulation verifies the universality of the fault voltage forward traveling wave energy rate of change feature and the exponential coefficient feature in both grid-connected and islanded operation modes. The study of the characteristics of the traveling wave before fault voltage lays a theoretical foundation for the application of the traveling wave principle in the field of protection.

(2) An adaptive start-up criterion based on the deviation of the energy change rate of the fault voltage forward wave is proposed. Based on the characteristics of the energy rate of change of the fault voltage forward traveling wave, a sliding time window algorithm is introduced to construct the energy rate of change deviation, so as to judge whether a fault occurs or not. The proposed criterion can effectively reflect the sudden change of the energy change rate of the fault voltage forward traveling wave and realize the adaptive adjustment threshold, which solves the problems that the traditional start-up criterion can't start up properly and the adaptability is not good enough. The feasibility of the proposed adaptive start-up criterion under different fault locations, fault types and operation modes are verified by simulation.

(3) A protection action criterion based on the exponential coefficient characteristic of fault voltage forward traveling wave and the Levenberg-Marquardt (L-M) method is constructed, and a microgrid protection method based on fault voltage forward traveling wave characteristics is proposed. The exponential coefficients are identified by fitting and combining the fault voltage forward wave by the L-M method, and the maximum value of the exponential coefficients is obtained. The fault voltage forward traveling wave exponential coefficient characteristics are utilized for fault identification, and the protection action criterion is constructed. Combined with the start-up criterion based on the deviation of the energy change rate of the fault voltage forward traveling wave, a protection method based on the fault voltage forward traveling wave characteristics is constructed, and a specific protection process is designed. The proposed method no longer relies on traveling wave head information, and can effectively avoid the problem of protection refusal or false action caused by the difficulty of traveling wave head extraction. Simulation results show that the microgrid protection method based on fault voltage forward traveling wave characteristics can accurately identify faults under the conditions of different fault locations, different fault types and different operation modes, thus verifying the feasibility and reliability of the proposed method.