基于稀疏量测的海上风电场集电线路故障选线方法研究
RESEARCH ON FAULT LINE SELECTION METHOD FOR OFFSHORE WIND FARM COLLECTING LINES BASED ON SPARSE MEASUREMENT
王晓东 1吴家豪 1高兴 1刘颖明1
作者信息
- 1. 沈阳工业大学电气工程学院,沈阳 110870
- 折叠
摘要
针对海上风电场多分支集电线路故障定位大都依赖于多测点的问题,提出一种基于卷积神经网络(CNN)的集电线路故障选线方法,基于稀疏量测利用局部连接实现集电线路故障选线.该方法以少量节点电流信号作为特征量,建立以稀疏样本的CNN初始网络损失最小为目标的量测位置优化模型,利用离散二进制粒子群(BPSO)算法进行模型求解得出最优量测位置.算例分析表明,所提方法可在稀疏量测下以较高精度实现故障选线,对采样频率要求较低,不受故障起始角、故障电阻、故障位置等因素的影响,且对量测噪声具有较好的鲁棒性.
Abstract
To solve the problem that the fault location of multi-branch collecting lines of offshore wind farm mostly depends on multiple measurement points,a method of fault line selection of collecting lines based on convolutional neural network(CNN)is proposed,which uses local connection based on sparse measurement to realize fault line selection of collector lines.In this method,a small number of node current signals are taken as the characteristic quantity,and a measurement position optimization model is established with the goal of minimizing the initial CNN network loss of sparse samples.The binary particle swarm optimization(BPSO)algorithm is used to solve the model and obtain the optimal measurement position.The example analysis shows that the proposed method can achieve fault line selection with high accuracy under sparse measurements,with low sampling frequency requirements,and is not affected by factors such as fault starting angle,fault resistance,and fault location.It also has good robustness to measurement noise.
关键词
海上风电场/集电线路/卷积神经网络/离散二进制粒子群优化算法/故障选线/量测位置Key words
offshore wind farms/collector lines/convolutional neural network/binary particle swarm optimization algorithm/fault line selection/measurement position引用本文复制引用
出版年
2024
NETL
NSTL
万方数据