摘要
为更加高效智能地识别出驾驶员值班行为特征,提出一种基于Wi-Fi信道状态信息(CSI)的驾驶员值班行为识别方法.在分析驾驶员值班过程中不同行为对CSI时域和频域影响的基础上,通过提取CSI中表征不同值班行为对应的速度特征,构建包括驾驶员静止、行走、原地操作活动三种不同值班状态的隐马尔可夫模型(HMM),利用HMM对人体行为特征序列进行判别,从而实现驾驶员值班行为状态的分类.实船测试表明,该方法具有较高的准确性和鲁棒性,识别不同值班状态的平均准确度可达90.3%,可为智能航运监管、船舶航行安全提供帮助.
In order to identify the characteristics of OOW’s watch-keeping behavior more efficiently and intelligently,a method of watch-keeping behavior recognition based on Wi-Fi channel state information(CSI)was proposed.Based on the analysis of the impact of different behaviors on the CSI time and frequency domain during OOW’s watch-keeping process,by extracting the speed characteristics corresponding to different behaviors from CSI,a hidden Markov model(HMM)including three different duty states of OOW as stationary,walking and in situ operation was constructed,and HMM was used to distinguish the human behavior feature sequence to realize the classification of OOW on duty behavior.The test on board proves that the proposed method has higher accuracy and robustness,and the average identifying accuracy of different duty states can reach 90.3%,which can provide help for intelligent shipping supervision and ship navigation safety.
作者
陈嘉鸣
刘克中
陈默子
马杰
曾旭明
CHEN Jia-ming;LIU Ke-zhong;CHEN Mo-zi;MA Jie;ZENG Xu-ming(School of Navigation,Wuhan University of Technology,Wuhan 430063,China;Hubei Provincial Key Laboratory of Inland Navigation Technology,Wuhan University of Technology,Wuhan 430063,China)
出处
《大连海事大学学报》
CAS
CSCD
北大核心
2020年第3期68-75,共8页
Journal of Dalian Maritime University
基金
国家自然科学基金资助项目(51979216)
湖北省技术创新专项重大项目(2017AAA120)。