有源配电网逆变器SVM改进及保护灵敏度增益

SVM Improvement and Protection Sensitivity Gain for Active Distribution Network Inverters

中文关键词:有源配电网  逆变器  负序电流  故障识别  快速SVM策略

keywords:active distribution network  inverter  negative sequence current  fault identification  fast SVM strategy

中文摘要:如何解决电网故障快速识别与选择性保护的问题，对系统中逆变器故障响应能力提出了全新要求。提出一种面向电网稳定运行的有源配电网逆变器改进空间矢量调制（space vector modulation，SVM）策略。基于SVM与正弦脉冲宽度调制（sinusoidal pulse width modulation，SPWM）的数理等效性，通过重构调制波零序分量生成机制，在保证电网稳定性的同时，重点优化故障的负序电流动态特性：稳态运行时严格抑制负序电流，避免保护误动；故障暂态阶段动态增强负序电流输出，使其快速满足保护检测，提升故障特征辨识度。在不对称故障中，负序电流不仅帮助识别故障特征，还能有效定位故障源。改进策略在满足继电保护快速性与选择性要求的同时，显著降低了计算负担。实验结果表明，该方法可缩短保护动作时间，提高电网故障检测的可靠性，有效增强有源配电网逆变器与继保装置在电网故障识别中的协调与配合能力。

Abstract:Addressing the challenge of rapid fault identification and selective protection has imposed entirely new demands on the fault-response capability of inverters within the system. For this end， this paper proposes an improved space vector modulation （SVM） strategy for active distribution network inverters to ensure the stable operation of the power grid. Leveraging the mathematical equivalence between SVM and sinusoidal pulse width modulation （SPWM），the zero-sequence component generation mechanism is reconstructed. The method suppresses negative-sequence currents in steady-state operation to prevent false tripping， while dynamically boosting their output during fault transients to quickly meet protection detection thresholds and improve fault feature discrimination. Under asymmetric faults， negative-sequence currents not only help fault characterization but also support accurate fault location. The proposed strategy satisfies the speed and selectivity requirements of relay protection with reduced computational burden. Experimental results verify that it shortens protection action time， enhances the reliability of grid fault detection， and improves the coordination and cooperation between active distribution network inverters and relay protection devices in grid fault identification， thereby supporting the stable operation of active distribution networks.