|国家科技期刊平台
首页|期刊导航|现代电子技术|基于混合深度学习的电动汽车充放电系统通信协议辨识方法研究

基于混合深度学习的电动汽车充放电系统通信协议辨识方法研究OA北大核心CSTPCD

Communication protocol identification method based on hybrid deep learning for electric vehicle charging and discharging system

中文摘要英文摘要

电动汽车充放电系统是新能源汽车安全充放电的重要保障.为了解决电动汽车充放电系统的通信协议辨识问题,提出一种基于混合深度学习的电动汽车充放电系统通信协议辨识方法.该方法通过深度学习网络(DLN)提取通信协议数据特征,通过引入l1/2 范数提高深度学习网络的泛化能力,实现对电动汽车充放电系统通信协议的高精度辨识.仿真结果表明,混合深度学习网络对电动汽车充放电系统不同协议的辨识总体准确率达到了97.68%.因此,可以得出基于混合深度学习的电动车充放电系统通信协议辨识方法具有一定的有效性.

The electric vehicle(EV)charging and discharging system is an important guarantee for the safe charging and discharging of the new energy vehicles(NEVs),so a communication protocol identification method based on hybrid deep learning for the EV charging and discharging system is proposed to cope with the communication protocol identification of the system.In this method,the communication protocol data features are extracted by the deep learning network(DLN),and the generalization ability of this network is improved by introducing norm l1/2,so as to achieve high-precision identification of the communication protocols of the EV charging and discharging system.The simulation results show that the overall accuracy(OA)of the hybrid deep learning network for the identification of different protocols of the EV charging and discharging system reaches 97.68%,which indicates the effectiveness of the communication protocol identification method for the EV charging and discharging system based on hybrid deep learning.

吕晓荣;惠琪;许子旻

国电南瑞南京控制系统有限公司,江苏 南京 210000南京邮电大学 自动化学院/人工智能学院,江苏 南京 210000

电子信息工程

电动汽车充放电协议辨识深度学习网络l1/2范数评估指标泛化能力

EV charging and discharging systemprotocol identificationDLNnorm l1/2evaluation indicatorgeneralization ability

《现代电子技术》 2024 (017)

41-46 / 6

国电南瑞南京控制系统有限公司科技项目:满足国际化要求的新一代充电技术研究及关键设备研制(524609230048)

10.16652/j.issn.1004-373x.2024.17.007

评论