| Title | Develop a risk data system for bicyclists riding on the roads |
| Dept | Transportation Safety Division |
| Year | 2013 |
| Month | |
| Price | |
| Summary | The pupose of this study was to build a system for traffic risk data collection, processing, and analyzing. Firstly, this study developed an annotation and reduction system to extract and code information relevant to naturalistic cycling data based on the data collected in a previous IOT study. Secondly, based on a prototype of instrumented bicycle proposed in the previous IOT study, some on-board instruments were modified as necessary. Finally, the data collected by the instrumented bicycle being ridden in the real traffic were reduced and annotated for risk analysis. An applicable risk data system for bicyclists riding on the roads was built in the present study.
The system included cycling data collecting, data pre-processing and merging into a cycling behavior database, traffic and environment conditions annotating, and three-staged traffic event analysis. An assistant computer sofeware for data processing was also developed. Additionally, a data reduction and an annotation method were proposed. This method used six parameters’ thresholds, including lateral acceleration, lognitudial acceleration, wheel angle, left-side ultrasonic sensed distance, right-side ultrasonic sensed distance, and lognitudial acceleration from speed differential, to set up three trigar rules. Such rules can reduce the numberous raw data to 20% with meaningful data. A combination procedure further reduced the numbers of data to 2% for data annotation. In the present study, some cases analysis have demonstrated the applicabilities of data treated through these procedures. |
| Post date | 2013/12/05 |
| Hashtags |
View count:
113