| Nowadays, much research is carried out on the travelling characteristics of heavy vehicles with the increasing numbers of heavy vehicle on the road, but little research is done on the turning features, further traveling influence analyzing and simulation of the heavy vehicles in microscopic traffic flow simulation area.In the first part of this paper, a research review is introduced, including recent research on car-following characteristics,overtaking and the psychological features of the drivers, and turning feature analysis of the heavy trailer, like drifting, overturning and side-slide behavior analysis and computer simulation research; next, cellular automaton theory is described, and cellular automaton models of single-cell, multiple-cell, multiple-lane and urban traffic flow are summarized and analyzed in this paper.Next, by analyzing the stressing situation of the heavy trailer turning, the driving track of the heavy trailer in normal and drifting driving conditions is deduced. A driver decision-making model is established, and we design the state awareness vector and component value constraints, which can help the driver to make corresponding action decisions, such as acceleration, deceleration and turning. Next, we use the VBOX to collect and analyze the turning data of the heavy trailer vehicles and ordinary passenger cars in Tianjin Port road, using which to value the state awareness vector. Then, the heavy trailer turning model is established based on cellular automation, in this model, the vehicle driving track and the driving decision can be described dynamically, and we also can get the simulation result of next action to the driving track and the traffic around.Finally, we design many groups of curve driving simulation experiments under certain circumstance, in which traffic simulation data and space-time diagrams can be got, The simulation results show that the curve design parameters of road with a high proportion of heavy trailer can be tested and corrected in this model, including lane width, ultrahigh and curve radius, by which the effectiveness and feasibility of this model can be verified. |
