WiDriver:一种基于WiFi的驾驶员情绪波动识别框架  

WiDriver:a WiFi-based Driver Emotion Fluctuation Recognition Framework

作　　者：周程宁 王青山[1] 王琦[1] 沈德伟 ZHOU Cheng-ning;WANG Qing-shan;WANG Qi;SHEN De-wei(School of Mathematics,Hefei University of Technology,Hefei 230031,China)

机构地区：[1]合肥工业大学数学学院,合肥230031

出　　处：《小型微型计算机系统》2022年第10期2137-2142,共6页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61571179,61401144)资助.

摘　　要：驾驶过程中因为情绪波动过大而造成的驾驶状态失常现象,如路怒症等,是导致道路交通事故的重要因素之一,甚至可能造成严重的人员伤亡.现有的情绪波动识别检测工作主要是基于视觉和生物信号传感器的检测手段.然而,基于视觉的方法具有视觉阻塞或失真问题,基于生物信号的方法具有侵入性、隐私侵犯等问题,且其使用的设备也会带来不便或额外成本.本文提出一种新的基于WiFi信号的情绪波动识别框架WiDriver,以克服现有方法的不足.WiDriver首先通过菲涅耳区设计天线位置区域以达到最佳信号采集效果.其次,通过收集驾驶员油门与刹车动作的信道状态信息(Channel State Information)进行情绪识别系数计算,通过基于识别系数与LSTM的情绪判别器进行情绪波动识别.实验将WiDriver部署在商业WiFi基础设施中,并评估其在真实驾驶环境中的性能.实验结果表明WiDriver在真实场景中的平均识别率达到83.9%.Driving disorders caused by excessive emotional fluctuationsin the process of driving,such as road rage,are one of the important factors leading to road traffic accidents,which may even cause serious casualties.The existing detection methods of emotion fluctuation are mainly based on visual and biological signal sensors.However,the visual-based approach has problems of visual blocking or distortion,and the biosignal-based approach has problems of intrusion,privacy invasion,and the use of equipment may also bring inconvenience or additional cost.In this paper,WiDriver,a WiFi signal-based emotional fluctuation recognition framework,is proposed to overcome the shortcomings of existing methods.WiDriver first designs the antenna location area through the Fresnel region to achieve the best signal acquisition effect.Secondly,the emotion recognition coefficient was calculated by collecting the channel State Information of the driver′s throttle andbrake actions,and emotion fluctuation was recognized by the emotion discriminator based on the recognition coefficient and LSTM.The experiment deployed WiDriver in a commercial WiFi infrastructure and evaluated its performance in a real-world driving environment.The experimental results show that the average recognition rate of WiDriver in the real scene is 83.9%.

关 键 词：驾驶员情绪波动 WIFI 信道状态信息 情绪系数 LSTM 

分 类 号：TP302[自动化与计算机技术—计算机系统结构]