Dynamic Complexity Of Driving:Modelling And Experimental Study

This research’s main aim is to examine the effect of dynamic driving environmentupon car drivers using dynamic driving complexity. To achieve the aim, four steps havebeen taken. Namely, finding key factors of dynamic driving complexity, buildingcomplexity conceptual model, proposing a method which is based on time-to-collisionto measure dynamic driving complexity, and assessing the measurement’ effectiveness.A simulator experiment has been carried out to evaluate the measurement ofdynamic driving complexity. There were four tasks which had different complexitylevel. Overall measures of driving such as driver’s speed and position were collected,together with subjective measures of driver perceived driving complexity and workload.One of the research’s findings is that there was a strong link betweentime-to-collision and driver’s brake performance. The probability of braking vehiclewas significantly higher when the time-to-collision fell into interval (0,10](p <0.001)at the0.05level. The severity of brake was a function of time-to-collision, y=15.631x-0.385, R2=0.690and p <0.001.Another finding is that braking vehicle partly resulted from instantaneous dynamicdriving complexity. For task2, the correlation coefficient was0.5with confidence level0.05. For task3and task4, the correlation coefficient was0.4and0.45.The third finding is that dynamic driving complexity was significantly correlatedwith subjective complexity (spearman correlation coefficient=0.522, p <0.001),subjective workload (spearman correlation coefficient=0.502, p <0.001), total numberof brake (spearman correlation coefficient=0.612, p <0.001), and the sum of brakeamplitude (spearman correlation coefficient=0.718, p <0.001) at the0.05level.The conclusion of the research is dynamic driving complexity is an effectivemeasurement of the effect of dynamic driving environment on vehicles driver.