Statistics Of Multi-lane Random Traffic Flow And Research On Equivalent Load Model For Long-span Bridge

As the increasing of bridge span, the number and speed of the vehicles on the bridge and the appearance of heavy and over-heavy vehicles, the random traffic flow on the bridge becomes more and more complex. Due to the random traffic flow load of long-span Bridges have strong randomness, the traditional research method that assuming the vehicular gap and the speed of vehicles is constant has great limitations and difficult to have a accurate analysis of the characteristics of random traffic load. With the random traffic data from the toll station of the second Yangtze River Bridge of Nanjing and the testing traffic flow data of the bridge, the distribution pattern and parameters of different vehicle type’s weight was ascertained, and the proportion of vehicle type and vehicle average arrival rate of each lane at different traffic service level were also determined. Based on the cellular automata model(CA model), begin with the cellular automata model of single lane in the field of traffic control and combine with the proportion of vehicle type and vehicle average arrival rate of each lane to build a mathematical model of multi-lane random traffic flow. And the multi-lane random traffic flow simulation program was developed based on Matlab program to simulate the random traffic flow of the second Yangtze River Bridge of Nanjing. Finally, the characteristic of traffic flow in the process of random traffic operation was studied, the main content is depicted as follows:(1) The driving rules of the vehicles on bridges were researched. It includes the principle of acceleration and deceleration and lane changing rules. It were determined that how to change lane when vehicles need changing lanes and the probability of lane changing. The mathematical model of multi-lane random traffic flow was was presented based on the operation rules of traffic flow and the cellular automata model.(2) The multi-lane random traffic flow simulation program of long-span bridge was developed based on the mathematical model of random traffic flow and Matlab program to simulate the random traffic flow of the second Yangtze River Bridge of Nanjing. The output interface of simulation program was designed by using the functions of interface design and handle graphics, and the visualization of random traffic flow simulation was realized. The implementation of the program can provide the support for the generation of random traffic flow load.(3) A large number of characteristic parameters of random traffic flow such as the proportion of vehicle type, vehicle average arrival rate of each lane at different traffic service level and vehicle weight data of different vehicle types were obtained from the toll station of the second Yangtze River Bridge of Nanjing and through the direct observation of the random traffic flow. Chi-square goodness of fit tests are performed on the vehicle average arrival rate of this traffic survey, it discovers that vehicle average arrival rate at different traffic service level basically follow poissonian distribution.(4) The random traffic flow of long-span bridge was simulated based on the testing traffic flow data of the bridge and cellular automata model theory. It proves the accuracy and reasonability of simulation program according to the comparison between simulative traffic flow data and measured one. The running characteristics of traffic flow at different traffic service level were studied. Moreover, the regularity and characteristics of traffic, speed and vehicular gap of random traffic flow were futher analyze. It shows that vehicular gap at three traffic service levels basically obey the lognormal distribution. The corresponding logarithmic normal distribution curves were fitted, and it was proved based on kolmogorov-smimov goodness-of-fit test.(5) The space distribution of vehicles the weight, speed of corresponding vehicles were obtained by simulating random traffic flow. The equivalent dynamic wheel load (EDWL) approach was introduced to transform random traffic flow location to random traffic flow load. The mathod can simplify interaction between vehicles and bridge. The load time-history curve of cross section of bridge was obtained.