Post-earthquake rescue missions are full of challenges due to the unstable structure of ruins and successive aftershocks.Most of the current rescue robots lack the ability to interact with environments,leading to low ...Post-earthquake rescue missions are full of challenges due to the unstable structure of ruins and successive aftershocks.Most of the current rescue robots lack the ability to interact with environments,leading to low rescue efficiency.The multimodal electronic skin(e-skin)proposed not only reproduces the pressure,temperature,and humidity sensing capabilities of natural skin but also develops sensing functions beyond it—perceiving object proximity and NO2 gas.Its multilayer stacked structure based on Ecoflex and organohydrogel endows the e-skin with mechanical properties similar to natural skin.Rescue robots integrated with multimodal e-skin and artificial intelligence(AI)algorithms show strong environmental perception capabilities and can accurately distinguish objects and identify human limbs through grasping,laying the foundation for automated post-earthquake rescue.Besides,the combination of e-skin and NO2 wireless alarm circuits allows robots to sense toxic gases in the environment in real time,thereby adopting appropriate measures to protect trapped people from the toxic environment.Multimodal e-skin powered by AI algorithms and hardware circuits exhibits powerful environmental perception and information processing capabilities,which,as an interface for interaction with the physical world,dramatically expands intelligent robots’application scenarios.展开更多
针对当前无线通信网络节点攻击入侵告警存在准确性差、告警响应速度慢的问题,引入改进长短期记忆网络(Long Short Term Memory,LSTM),开展无线通信网络节点攻击入侵告警算法研究。首先,收集无线通信网络数据,提取受损节点特征;其次,利...针对当前无线通信网络节点攻击入侵告警存在准确性差、告警响应速度慢的问题,引入改进长短期记忆网络(Long Short Term Memory,LSTM),开展无线通信网络节点攻击入侵告警算法研究。首先,收集无线通信网络数据,提取受损节点特征;其次,利用改进LSTM构建节点攻击入侵检测模型;最后,结合模型输出,对攻击入侵行为告警,并完成告警信息融合。实验结果表明,新的告警算法可以实现对所有无线通信网络节点攻击入侵行为的准确告警,且响应速度显著加快。展开更多
基金supports from the National Natural Science Foundation of China(61801525)the independent fund of the State Key Laboratory of Optoelectronic Materials and Technologies(Sun Yat-sen University)under grant No.OEMT-2022-ZRC-05+3 种基金the Opening Project of State Key Laboratory of Polymer Materials Engineering(Sichuan University)(Grant No.sklpme2023-3-5))the Foundation of the state key Laboratory of Transducer Technology(No.SKT2301),Shenzhen Science and Technology Program(JCYJ20220530161809020&JCYJ20220818100415033)the Young Top Talent of Fujian Young Eagle Program of Fujian Province and Natural Science Foundation of Fujian Province(2023J02013)National Key R&D Program of China(2022YFB2802051).
文摘Post-earthquake rescue missions are full of challenges due to the unstable structure of ruins and successive aftershocks.Most of the current rescue robots lack the ability to interact with environments,leading to low rescue efficiency.The multimodal electronic skin(e-skin)proposed not only reproduces the pressure,temperature,and humidity sensing capabilities of natural skin but also develops sensing functions beyond it—perceiving object proximity and NO2 gas.Its multilayer stacked structure based on Ecoflex and organohydrogel endows the e-skin with mechanical properties similar to natural skin.Rescue robots integrated with multimodal e-skin and artificial intelligence(AI)algorithms show strong environmental perception capabilities and can accurately distinguish objects and identify human limbs through grasping,laying the foundation for automated post-earthquake rescue.Besides,the combination of e-skin and NO2 wireless alarm circuits allows robots to sense toxic gases in the environment in real time,thereby adopting appropriate measures to protect trapped people from the toxic environment.Multimodal e-skin powered by AI algorithms and hardware circuits exhibits powerful environmental perception and information processing capabilities,which,as an interface for interaction with the physical world,dramatically expands intelligent robots’application scenarios.
文摘针对当前无线通信网络节点攻击入侵告警存在准确性差、告警响应速度慢的问题,引入改进长短期记忆网络(Long Short Term Memory,LSTM),开展无线通信网络节点攻击入侵告警算法研究。首先,收集无线通信网络数据,提取受损节点特征;其次,利用改进LSTM构建节点攻击入侵检测模型;最后,结合模型输出,对攻击入侵行为告警,并完成告警信息融合。实验结果表明,新的告警算法可以实现对所有无线通信网络节点攻击入侵行为的准确告警,且响应速度显著加快。
