车载信息系统交互方式对驾驶行为的影响研究

Influence of in-vehicle information system interaction modes on driving behavior

车载信息系统手动和语音交互方式给驾驶人带来新型交互体验的同时,也不可避免地引起了驾驶分心问题。为明确真实道路条件下车载信息系统手动和语音交互方式对驾驶行为的影响,招募20名被试佩戴Dikablis头戴式眼动仪,开展了2(语音和手动)×3(电台调频、导航设置和音乐播放)城市道路实车驾驶试验。两因素重复测量方差分析显示,交互方式和操作类型对驾驶性能、扫视行为及工作负荷具有显著影响。事后比较检验表明,导航设置任务对驾驶分心的影响最大,音乐播放次之,电台调频最小。相比于基线任务(仅驾驶),语音和手动次任务的介入导致车速下降了约15.2%和25.1%,车道偏离预警次数增加了65.4%和90.9%。同时,相比于语音交互,手动交互次任务导致平均扫视时间、扫视次数和总扫视时间显著增加;此外,描述性统计显示手动交互的负荷大于语音交互。研究揭示了车载信息系统交互方式对驾驶行为的影响机理,为车载信息系统新型交互技术开发和应用提供了实证依据。

The manual and voice-based interaction with the in-vehicle information system(IVIS)not only bring new interactive experience to the driver but also inevitably cause the problem of driving distraction.To clarify the influence of manual and voice-based interaction with IVIS on driving behavior under real road conditions,twenty participants from China were recruited wearing Dikablis head-mounted eye tracker to carry out driving experiments in urban road environment with two-way six lanes.Among them,the within-subject experimental design of 2(manual and voice-based)×3(frequency modulation,navigation,and music)was adopted.The two-way repeated-measures ANOVA showed that the interaction mode and operation type of IVIS had a significant influence on driving performance,glance behavior,and NASA-TLX.However,no insignificant interaction was found.A post-hoc test showed that navigation had the greatest impact on driving distraction,followed by music and FM.The intervention of voice and manual-based secondary tasks resulted in an average speed decrease of about 15.2%and 25.1%.Similarly,the number of lane departure warnings increased by 65.4%and 90.9%compared to the baseline task(driving only),respectively.Besides,compared with voice-based interaction,manual-based interaction with an in-vehicle information system resulted in a significant increase in total glance time and number of glances,but no significant difference in average glance time.Furthermore,descriptive statistics showed that the subjective load of manual-based interaction is greater than that of voice-based interaction.That was to say,under the condition of manual-based interaction while driving,drivers felt more driving loads and commonly take the strategy to lengthen the average glance time and increase the number of glances to obtain more secondary-task related in-vehicle information.This study revealed the influence mechanism of interaction modes of IVIS on driving behavior,which provided an empirical basis for the development and application of new interactive technology of IVIS.