Modeling And Simulation With CPN For Real-time Traffic Control System Based On ITS Architecture

Urban traffic signal control system has been a hot topic of research by domestic andforeign scholars. Rational and efficient urban traffic signal control system does not onlyhelp to improve the existing transportation system performance, but also effectivelyimprove the living environment of the city and produce considerable social benefits for thecity.According to complex engineering system benchmarks research areas: the process,the architecture frameworks (AFs), the methodologies, the modeling and evaluation.Based on the analysis of current urban traffic signal control system’s research situation anddevelopment trends at home and abroad, a number of key technologies of urban trafficsignal control are systematically studied and discussed in this paper, especially research onvihicle detection and classification system, model of urban traffic flow based on ITSarchitecture and Hierarchical Colored Petri Nets(HCPN), and a new real-time adaptiveurban traffic signal control algorithm.The paper mainly covers the following areas:Firstly, a novel vehicle detection and classification system based on anisotropicmagnetoresistive (AMR) sensors and support vector machine (SVM) algorithm ispresented. The AMR sensors detect the change of earth magnetic field which will bedisturbed differently by different types of passing traffic vehicle. With the featuresextracted from the disturbance data of earth magnetic field, SVM classifier will be trainedand tested to classify the vehicle type. The results of our experiments indicated thatvehicle classification based on AMR sensors and SVM algorithm is effective and efficient.In addition, using AMR vehicle detection sensors as sensor network nodes, networkedsensors vehicle detection system based on the low cost and high-speed rs485serial bus ispresented.Secondly, after researching on urban transportation system architecture and trafficflow theory, an urban ITS architecture and a hierarchical modeling approach for urbantraffic signal control system are established based on Colored Petri Nets, whicharchitecture inlcudes three levels: the top level-operational view, the middle-systemview, and the underlying-technical standard view. According to the characteristics ofurban traffic flow, a hierarchical structure model of urban traffic flow is presented, whichmodel consists of three levels, traffic network level, network conponents level and logic control level, and implements the seamless connection between traffic flow and logiccontroller. Based on hierarchical Colored Petri Nets theory and actual road parameters,models of every level are established using top-down modeling approach. Afterestablishing the hierarchical Colored Petri Nets model of a real traffic network forexample, introduced the design approach of Colored Petri Nets model, argued that themodel has good extensibility, and validated that the hierarchical structure model will makeus better to achieve remarkable results in complex problem solving tasks on traffic flowmodeling.Finally, according to the analysis of urban traffic flow model and mainly taking thecapability of real-time and adaptive control into account, traffic signal control algorithmsfor isolated intersection and area traffic are presented respectively. Based on the trafficflow characteristics of an isolated intersection and precdiction for short-term traffic flow,aiming to minimize the vehicle delay, a rule-based traffic signal control algorithm ispresented. After analyzing the relationship of traffic flow between adjacent intersections, arule-based coordination control algorithm is presented with traffic volume in the linkbetween adjacent intersections as the coordination variable. The property of the trafficflow model with HCPN is analyze, and then through the simulation, the evaluatingindicator is created to evaluate the control strategy. The consequence of the simulationindicates the real-time adaptive traffic signal strategy achieves more effective control.