基于混合量测的谐波源定位及其发射水平评估研究

Harmonic state estimation proposed by most literatures is totally based on PMU measurements and too expensive to achieve. On the other hand, Chinese power companies have constructed a great number of Power Quality Monitoring systems and got a lot of first-hand harmonic datum, but lacked of effective analysis tools. In order to solve this problem, the project proposed a harmonic state estimation method based on SCADA and PMU mixed measurments. To overcome the inevitable measurement error, synchronization error and parameter error, a nonlinear model robust least squares estimation algorithm, which can achieve well-balance between the convergence speed and the robust performance, is introduced during the harmonic state estimation. In order to further improve the estimation accuracy, Key technologies such as multi-objective Optimal Measurement Placement Method and augmented matrix harmonic network parameters identification are researed in details. What's more, based on the reference impedance method and IEC definition on user's harmonic voltage emission level, this project proposes a more reasonalbe vector contribution index and corresponding evaluation algorithm to distinguish each nonlinear users' harmonic pollution responsibility, which is compatible with the current harmonic GB management indicators. Finally, a new economical and practical engineering strategy for regional power grid harmonic source localization and quantitative analysis was achieved by combining the system oriented harmonic state estimation methods with the equivalent circuit model based quantitative evaluation methods. The research results will be a breakthrough in Chinese regional power grid harmonic source localization and quantitative analysis techniques, and will promote the introduction of the harmonic source incentive management regulations.

本项目提出一种能够综合利用同步量测值和异步量测值的混合谐波状态估计方法，解决完全基于PMU同步量测技术的评估方案价格过于昂贵，而已有电能质量监测网获取的大量原始数据缺乏有效分析手段的问题。通过引入非线性模型抗差最小二乘估计算法来克服不可避免的量测误差、同步误差和参数误差，同时兼顾算法的效率和抗差性能。采用多目标优化量测配置、增广矩阵辨识谐波网络参数等方法进一步提高估计精度。并在IEC关于用户谐波电压发射水平定义的基础上，基于参考阻抗法，构建出一套与现行谐波国标管理指标兼容的、能够更合理区分用户责任的基于"矢量贡献"的多谐波源污染责任分解评价指标及其计算方法。将面向系统的谐波状态估计方法与基于等效电路模型的谐波源定量评估方法结合起来，提出一套经济实用的地区电网谐波源定位和定量分析的工程实施方案。研究成果将会使我国地区电网谐波源定位及定量分析技术取得突破性进展，推动谐波源奖惩管理法规的出台。

在谐波污染问题日益严重的今天，研究并提出实用的能够依据当前电网谐波监测数据对谐波源进行定位和定量评价的指标和方法，使得电力行业对谐波污染的技术监督能够像功率因数奖惩管理一样简单有效，具有非常重要的意义。鉴于此，本项目的主要研究内容定为：①基于参考阻抗法的谐波源定位及评价指标；②电能质量数据交换格式下的谐波责任划分方法；③多点谐波源责任划分方法；④利用分布式电源抑制配网谐波污染的方法。通过上述研究，本项目得到的结论是：①参考阻抗法中负载侧参考阻抗应设为无穷大、系统侧参考阻抗一般应设为最大运行方式下的短路阻抗，并按用户谐波电压的矢量贡献来评价其谐波污染责任，这样可确保任意工况下对用户谐波责任划分的公平合理性；②可根据公共点谐波电压和电流的有效值和系统参考阻抗估算出公共点谐波电压和电流的相位差，该估算法精度较高，能同时提供对估算误差的估计，并确保任何情况下都不会过高估计用户的谐波责任；③多点谐波源责任划分的关键是辨识出谐波源与关注母线之间的互阻抗，可通过数据选择得到单谐波源波动时刻，然后利用基于阻抗归一化的波动量法进行辨识，比直接应用最小二乘法来近似求解更精确可靠；④采用基于同步旋转坐标系的d-q解耦积分控制算法取代传统的比例控制算法，可有效改善基于电压检测的配网谐波抑制算法的谐波补偿效果，且算法的积分参数可根据闭环极点移动轨迹来优化确定。本项目的开展将会促进电力行业谐波监测管理水平的提高，使得通过经济杠杆促使用户自觉采用先进的谐波抑制装置或技术成为可能，为未来智能配电网充分利用用户的微源设备的谐波补偿能力净化电网环境扫除技术障碍。