【摘要】：基于光波導邊界傳輸特性和人工神經網絡,文中介紹了一種可同時監(jiān)測輸電線路絕緣子表面等值鹽密和灰密的污穢在線監(jiān)測裝置。在詳細分析光波導表面污層對全反射光束影響機理的基礎上,提出了光波導和光源設計方案。通過不同濕度、不同污穢等級、不同灰鹽比條件下的積污標定試驗,獲取了大量樣本數(shù)據,并依此建立和訓練了人工神經網絡模型。檢驗結果和運行實踐表明,利用該裝置及訓練后的內嵌神經網絡模型,可實現(xiàn)鹽密和灰密的同時監(jiān)測,且測試精度滿足相關標準要求。
[Abstract]:Based on the propagation characteristics of optical waveguide boundary and artificial neural network, an on-line pollution monitoring device is introduced, which can simultaneously monitor the equivalent salt density and grey density of insulator surface of transmission line. Based on the detailed analysis of the influence mechanism of the surface fouling layer on the total reflected beam, the design scheme of the optical waveguide and the light source is proposed. A large number of sample data were obtained through the calibration tests under different humidity, different pollution grades and different lime-salt ratios, and an artificial neural network model was established and trained accordingly. The test results and operation practice show that both the salt density and grey density can be monitored simultaneously by using the device and the embedded neural network model after training, and the precision of the test meets the requirements of relevant standards.
【作者單位】： 第二炮兵工程大學;西安金源電氣股份有限公司;
【基金】：國家自然科學基金資助項目(61102170)~~
【分類號】：TM855.2

【參考文獻】

相關期刊論文 前7條

1 李t熝，

本文編號：2315260