| 引用本文: | 肖舒严,王强钢,周念成.基于稳健加权总体最小二乘的变压器局部放电定位[J].电力自动化设备,2020,40(8): |
| XIAO Shuyan,WANG Qianggang,ZHOU Niancheng.Partial discharge location of transformer based on robust weighted total least square[J].Electric Power Automation Equipment,2020,40(8): |
|
| 摘要: |
| 到达时间差、声速与传感器位置坐标的测量误差将导致局部放电时差定位出现误差，通过研究发现变压器的内部结构及各部分的材质差异可能造成到达时间差以及声速的测量值可能远远偏离真实值，这种测量误差被称作粗差，但目前的局部放电定位算法未将数据测量误差纳入模型，使得现有定位算法在计及测量误差特别是粗差的情况下，定位误差较大难以满足工程要求。为此，分析了局部放电测量数据粗差的来源及现有定位算法的缺陷，提出了基于稳健加权总体最小二乘的局部放电定位算法，采用含误差变量模型推导迭代公式，运用权函数修改含粗差数据在迭代过程中的权重，能够有效抑制测量数据的随机误差、系统误差和粗差对定位结果的影响，能更为准确地定位局部放电，有助于制定针对性的检修策略，提高检修效率，缩短停电时间。 |
| 关键词: 变压器 局部放电 时差定位 总体最小二乘 稳健估计 |
| DOI：10.16081/j.epae.202007034 |
| 分类号:TM41 |
| 基金项目:国家自然科学基金资助项目(51577018)；重庆市基础科学与前沿技术研究专项资助项目(cstc2015jcyjBX0033) |
|
| Partial discharge location of transformer based on robust weighted total least square |
|
XIAO Shuyan1, WANG Qianggang2, ZHOU Niancheng2
|
|
1.Jiangbei Power Supply Company, State Grid Chongqing Electric Power Company, Chongqing 401120, China;2.State Key Laboratory of Power Transmission Equipment & System Security and New Technology, Chongqing University, Chongqing 400044, China
|
| Abstract: |
| The measurement errors of TDOA(Time Differences Of Arrival),sound speed and sensor position coordinate cause PD(Partial Discharge) location errors. It is found that the difference of the internal structure and material of each part of transformer may cause the TDOA and the measurement value of the sound speed far from the true values, which are called gross errors. However, the current partial discharge location algorithm does not include the data measurement error into the model, which makes the existing location algorithm difficult to meet the engineering requirements. Aiming at this problem, the source of gross errors in partial discharge measurement data and the defects of existing partial discharge location algorithms are analyzed, and a location algorithm based on robust weighted total least square is proposed, the iterative formula is derived by using the EIV(Errors In Variables) model, and the weight function is used to modify the weights of the data with gross error in the iterative process. The proposed algorithm can effectively restrain the influence of random error, systematic error and gross error of measurement data on the location result, locate partial discharge more accurately, help to make the corresponding maintenance strategy, improve the maintenance efficiency and shorten the time of power cut. |
| Key words: electric transformers partial discharge TDOA location TLS robust estimation |
