Study On Multi-alignment Factors To Drivers’ Workload In Freeway

Alignment is not only the most important indicators overall the freeway design, but also an important part of freeway. Designing the freeway alignment based on the landform could decrease the drivers’ workload and crash rate. However, when choosing the alignment parameters, the designers often ignored the drivers’ workload. This situation caused that the drivers’ workload was very high when they were driving on some sectors of freeway. And it was very dangerous for drivers. In this paper, the relationship between the drivers’ workload and physiological indicators was analyzed. Chose the drivers’ ECG to reflect the drivers’ workload. Combined the alignment parameters and ECG, models could be made between the drivers’ growth rate of heart rate and multi-alignment factors. The models were used to forecast the drivers’ workload.Firstly, the concept of the drivers’ workload has been introduced by analyzing on the driving process. To reflect the workload, the indicators were summarized and the growth rate of heart rate were chosen. The relationship between the alignment indexes and drivers’ workload has been analyzed, too. At the same time, the extrinsic and intrinsic performance of driver’ workload were analyzed.Then, designed the experiments according to the experimental objectives that collecting the drivers’ heart rate. Collected the alignment parameters of the experimental sections. Gained the drivers’ heart rate and other data after the experiment. By the data pre-processing, the drivers’ growth rate of heart rate could be obtained on the experimental sections.By using the grey correlation analysis method, we have clearly known that these seven alignment parameters were closely related to the drivers’ growth rate of heart rate. The models which were based on the artificial neural networks and support vector machine have been made and the best models have been selected in comparison with relative error. On the basis of drivers’ growth rate of heart rate which was predicted in the use of these models, the alignment has been evaluated.