Urban Road Traffic System Efficiency Index System And An Approach Of Multimodal Network Design

Under the pressure of rapid urbanization and motorization, many cities face a series of negative problem of traffic congestion, energy consumption, air pollution, etc. On the limited road and energy resources, developing the potential of the road traffic system and improving the efficiency of road, is one of the best way to solve the city traffic congestion, to achieve energy saving and emission reduction.To improve the efficiency of road traffic system, the first task is:analysis of the connotation of efficiency, the influencing factors on efficiency, analysis of which part of system construction and management can improve the efficiency of the system. However, different subject has the different concept of efficiency of the transportation system, which makes the definition of efficiency not unified. Therefore, it is difficult to objectively evaluate the efficiency level of system, also difficult to research the transportation planning with the target of optimization efficiency.Rational planning of city road network, various transport modes would be to their best advantages. However, in recent years, as the city road planning idea led to the motor vehicle, the slow traffic is not attached to it. The slow road system planning method only focus on the local regulation,but the overall planning method is only a qualitative concept. But it hasn’t concernd pedestrain and bicycle networks, or driveway with pedestrain and bicycle lanes network.This work is supported by grants from the National Basic Research Program of China (2012CB725400) and National Natural Science Foundation of China (51238008). The research can be summarized in two parts. The first part is the efficiency of road traffic system,the second part is the approach,models and algorithms of overall planning and design driveway with pedestrain and bicycle lanes network, and efficiency as optimization objective.First, the road traffic system consists of three parts:road network system, traffic system, management and control system. From the interpretation of the initial definition of "efficiency", to classifying various definitions related to transportation system "efficiency", the efficiency of road traffic system was defined, and summed up three kinds of meaning. This paper mainly discusses the definition of "relationship between the input and output of the road traffic system". The input and output of three components of the system is analysed,and production process of traffic the service products is explained.Second, traffic system efficiency evaluation index system is constructed by analysis input and output of the three components of system. Each index in the system is defined, quantified and graded. Using expert questionnaire, the index weight is calculated by the order relation method. The evaluation index system is applied by using fuzzy comprehensive evaluation.Third, studies how to select the optimal objective of network design problem.Takes the efficiency of traffic system as the optimization objective, ie, travelling time, vehicle fuel quantity and exhaust emissions. Analysis of China’s mixed traffic characteristics, and according to the characteristics, design the driveway, on-street bicycle lanes and pedestrian sidewalk in a system. The approach is served by a bi-level model. The upper model is the functions of the three objectives. The lower model is an aggregated combined model to forecast travel choices. Improved Boyce’s simultaneous model, it is extended to the mixed traffic and other commoner case. The model is formulated,verified,and solved by a diagonalization algorithm.Fourth, analyses the three objective and multi-objective value of the schemes and the mode share of each scheme. Although the design roads is a small proportion of the network, the difference of different network design scheme between travel demand and the objective value was correctly explained.Compared the lower model and independent traffic model, no additional cost coefficient model, without considering the additional cost model. The goodness-of-fit of the lower model is 0.659, significantly higher than other models.Fifth, discusses how to design heuristic algorithm. Select three kinds of intelligent optimization heuristic algorithm--ant system algorithm, particle swarm optimization algorithm, genetic algorithm to solve the bi-level model. How to set up parameter of the algorithm is discussed in detail. Each algorithm is trial run, and algorithms are improved according to trial results.Before improved, the frequency of find the optimal solution is very low. Increasing the number of groups is also not very effective.Bring in global convergence mechanism, GABS with some parameters finding the optimal solution increases to 75%. Hybrid PSO and GABS, the algorithm finding the optimal solution and convergence capability reaches 90%-95%.