微型电脑应用2026，Vol.42Issue(4)：96-100,5.
基于电网数字化审计安全的恶意URL检测
Malicious URL Detection Based on Power Grid Digital Audit Security
摘要
Abstract
Aiming at the poor detection effect of malicious uniform resource locator(URL)caused by complex URL features and multiple variants in power grid digital audit security,a malicious URL detection method of CSPDarknet-particle swarm optimi-zation-based bidirectional long and short-term memory(PBiLSTM)is proposed.A URL dynamic character library is construc-ted based on the security features of power grid digital audit,and a single character is mapped to a dense vector by word embed-ding.The semantic features of URL are extracted by CSPDarknet neural network,and it is input into BiLSTM network opti-mized by particle swarm optimization algorithm(PSO)to complete efficient detection.The experimental results show that the detection accuracy of the proposed method for malicious URL detection is 98.2％,which is better than other detection meth-ods.关键词
电网数字化审计/恶意URL/URL字符库/双向长短期记忆Key words
power grid digital audit/malicious URL/URL character library/BiLSTM分类
信息技术与安全科学引用本文复制引用
邬奕强,康晓燕..基于电网数字化审计安全的恶意URL检测[J].微型电脑应用,2026,42(4):96-100,5.基金项目
国网上海电力项目(SGSH0000SJWT2310328) （SGSH0000SJWT2310328）
