Bi-layer Programming Control Of Urban Area With Environmental Constraint

The some certain areas such as scenic area, campus, and commercial street have high population density. If these areas are only permitted the appropriate amount of traffic demand to go through, not only for reducing pollutant emissions, but for the health of outdoor people, it plays a positive role. In this dissertation, the macroscopic traffic flow model integrated with a vehicular emission model was adopted to estimate urban traffic emissions. The control strategy is sought to relieve the urban traffic congestion and pollutant emissions. The main contributions are summarized as follows.Firstly, collecting precise data described the real field area in Hangzhou of China, a microscopic simulation model was established. A procedure was proposed for traffic simulation model calibration and one case was presented with a microscopic simulation software, namely Aimsun. A statistical method named Latin Hypercube Design was used to generate a distribution of collections of parameter values form a multi-dimensional distribution. Secondary development in Aimsun I/O’s function was utilized to calibrate and validate automatically. Finally, the calibrated Aimsun was verified to have higher accuracy.Secondly, a bi-level programming model with environmental constraint was formulated. The upper level aimed to minimize the total time spent in the certain urban areas through the macroscopic traffic flow model. A typical cross and T-shaped intersection phase structure of timing setting considered the phase sequence exchange were presented as constraint of upper level. In the lower level, the goal was to maximize the vehicle permission release rate. In this work, vehicular emissions model was adopted to calculate the emissions, and an environmental constraint was introduced by considering the requirement of air quality.Finally, for the proposed model based on a specific environmental constraints, the optimal strategy for arterial and area coordinated control was designed. Real data simulation experimental results indicated that our method not only reduce the total travel time at a rate of approximately 10%, but lower the concentration of PM2.5 as well.