首页|期刊导航|光通信研究|一种OPGW多参量冰厚组合预测方法

一种OPGW多参量冰厚组合预测方法OA

An OPGW Multi-Parameter Ice Thickness Combination Prediction Method

中文摘要英文摘要

[目的]光纤复合架空地线(OPGW)是一种专为高压输电线路设计的复合光缆,在输电线路中有着广泛的应用.但 OP-GW 光缆的覆冰问题会严重影响输电线路的安全稳定运行.为提高 OPGW 光缆覆冰厚度预测的精度,文章提出了一种 OP-GW 多参量冰厚组合预测方法.[方法]文章提出了一种融合改进的 Circle 映射和黄金正弦算法的减法平均优化器(ICGSA-BO)优化反向传播(BP)神经网络.该方法首先通过变分模态分解(VMD)分别将布里渊光时域反射仪(BOTDR)设备和气象传感器采集的 OPGW 光缆覆冰数据进行分解,得到不同频率的模态分量,即本征模态函数(Intrinsic Mode Functions,IMFs),以降低原始序列的不稳定性.其次,构建 BP 神经网络,以均方误差(MSE)作为损失函数,利用改进的 Circle 映射与黄金正弦算法构建 ICGSABO 算法以优化 BP 神经网络的权重和阈值,提升了模型的全局搜索能力与收敛速度.最后,将各模态分量输入 ICGSABO-BP 模型获得覆冰厚度的预测值,并对各分量预测结果进行叠加重构.[结果]实验结果表明,文章所提多参量冰厚组合预测方法在 0~30 mm 的覆冰厚度预测范围内显著优于其他模型,预测精度(R2)达到 0.962 22,MSE 为 0.247 79,验证了所提模型的有效性.[结论]该冰厚预测模型能充分利用光缆的温度、应变数据和气象数据的优势,生成更精确的预测结果,对电网安全监测具有重要的应用价值.

[Objective]Optical Fiber Composite Overhead Ground Wire(OPGW)is a kind of composite fiber optic cable de-signed for high-voltage transmission lines,which has a wide range of applications in power transmission lines.However,the ice-covering problem of OPGW cables can seriously affect the safe and stable operation of transmission lines.In order to improve the accuracy of the prediction of the ice-covering thickness of OPGW fiber optic cable,this paper proposes a multi-parameter ice thickness prediction method for OPGW.[Methods]In this paper,we propose a Combining Improved Circle Mapping and Gold-en Sine Algorithm with Subtraction-Average-based Optimizer(ICGSABO)to optimize the parameter selection of Back Propaga-tion(BP)neural network.In this method,the OPGW fiber optic cable ice-covering data collected by Brillouin Optical Time-Do-main Reflectometer(BOTDR)equipment and meteorological sensors are firstly decomposed by Variational Modal Decomposition(VMD)algorithm.This process yields modal components at different frequencies,termed Intrinsic Mode Functions(IMFs),thereby reducing the instability of the original sequence.Secondly,the BP neural network is constructed with the Mean Square Er-ror(MSE)as the loss function.The ICGSABO algorithm is constructed using the improved Circle mapping and golden sine algo-rithm to optimize the weights and thresholds of the BP neural network to improve the global search capability and convergence speed of the model.Finally,each modal component is input into the ICGSABO-BP model to obtain the predicted value of ice cover thickness,and the prediction results of each component are superimposed and reconstructed.[Results]The experimental results show that the multicomponent ice thickness prediction method significantly outperforms other models in the prediction range of ice cover thickness from 0 to 30 mm,with a prediction accuracy(R2)of 0.962 22 and a MSE of 0.247 79,which verifies the effectiveness of the proposed model.[Conclusion]This ice thickness prediction model can take full advantage of the strain of fi-ber optic cables,temperature data and meteorological data to generate more accurate prediction results,which has an important ap-plication value for power grid security monitoring.

刘丹妮;沈思鹏;窦增;孙浩然;苏伟佳

国网吉林省电力有限公司信息通信公司,长春 130000长春理工大学高技术产业有限责任公司,长春 130022国网吉林省电力有限公司信息通信公司,长春 130000国网吉林省电力有限公司信息通信公司,长春 130000国网吉林省电力有限公司信息通信公司,长春 130000

信息技术与安全科学

光纤复合架空地线光缆多参量冰厚预测变分模态分解

OPGW optic cablemulti-parameterice thickness predictionVMD

《光通信研究》 2026 (4)

179-185,7

国网吉林省电力有限公司科技资助项目(2023-46)

10.13756/j.gtxyj.2026.240239

评论