Summary: | 博士 === 國立清華大學 === 工業工程與工程管理學系 === 97 === According to the official investigation in Taiwan, 2007, driver’s distraction, speeding, and short headway keeping result in more than one third proportions of accidents on the freeway driving and are harmful to transportation safety. With the development of advanced technologies, lots of techniques that can share out drivers’ workload have been applied to in-vehicle task assistance and helpful information provision in the recent day. For in-vehicle assistance, the “control-aided” adaptive cruise control (ACC) and the “safety-aided” collision warning system (CWS) are widely discussed for advanced vehicle control and safety system (AVCSS). The purpose of these technologies is to enhance driving safety, convenience, and efficiency originally. However, improper system and interface design may cause against the result instead, such as distracting and confusing drivers. And further, the integration of different systems should also bring about negative effects as a result of the conflict, resource competition, or mutual interference between systems.
The objective of this research is to find out safe and human-based parameter settings on ACC and interface design on CWS. Also, to construct the human driver model when interacting with assist systems. Bus driving environment is the focused issue because traffic accidents will be more terrible on buses and need to be concerned. Therefore, professional bus drivers and the bus driving simulator are introduced to this research. This research will progress via three phases, human driver model conceptualization, ACC/CWS improvement, and model realization/validation. In the first phase, main tasks are the experiment of time-gap parameter improvement for ACC and the interface evaluation for CWS. The construction of ACC and CWS is referred to the standard issued by SAE International. The safety is defined by driving-related measures, such as driver’s subjective evaluation and driving performance. Parameters and interface improvements are for model realization, and the empirical driving data are for model validation in the final phase. Secondly, the conceptual human driver model is going to be brought up, with using ACC as an example. This prototype will be validated in the next phase. Finally, the concept model will be simulated based on the prototype from the second phase, and the settings and results from the experiment of phase 1.
The contribution of this research includes the dimensions of designers, end-users, and the government. The experimental results are flexible for specialists’ reference and applicable for related studies of ACC/CWS on buses. Drivers can also drive with these advanced technologies more safely, conveniently, and efficiently. The findings on the interaction among systems are also helpful for recommendation of driving safety by the government. Besides, the validated model not only can be applied to further evaluation of ACC, but it can also be the reference and performance prediction tool for advanced assist systems.
|