Theory And Method Research Of Wide-area Traffic Signal Coordinated Control Based On Group Decision-making

Traffic signal control system only can make full use of immeasurable advantages from a system point of view by designing traffic signal coordinated control efficiently. In the area traffic signal coordinated control, every intersection has its own objective as well as obeys whole interests. Signal control scheme will be commonly determined by multiple intersections and it's a non-linear multi-objective group decision-making problem. Considering the defects in existing theory and method of traffic signal coordinated control, combining with development needs and technical characteristics in modern urban traffic control, this dissertation has researched and discussed in the following areas, including dynamic division theory of traffic control subareas, arterial group multi-objective decision-making, multi-intersection group decision-making control model, and group decision-making support system of area-wide coordinated control. The main scientific research and research results in this dissertation include:1. Analyzing the influences of distance and traffic volume between neighboring intersections, and signal timing parameters of connected intersections, neighboring intersection meeting degree, as a new traffic parameter, is put forward to quantitatively describe the correlation degree between two neighboring intersections. The corresponding calculation method and scientific rationality are proposed. After that, by determining boundary point of the subarea, setting constraint conditions and evaluation criteria of division scheme, the division strategy of substep with layers is achieved and a division model of traffic control subareas based on the min-max principle of adjacent intersection distance and the theory of meeting degree is established. The whole division process which is used to get the optimal scheme is proposed. Under this condition, the subarea division schemes are analyzed and evaluated with the intersection group clustering algorithm. An example is presented to test and verify the division method of traffic control subareas based on the theory of meeting degree.2. Analyzing the relationship between different objective function, the objective functions under crowd and non-crowd traffic states are determined. Introducing the satisfaction principle in group decision theory, evaluated satisfaction functions of multiple objectives are proposed. Then, an intersection multi-objective dynamic decision model is established and signal control decision rule is formulated, with the application of improved multi-objective mixed genetic algorithm to solve the optimal decision variable and generate the control scheme. In addition, by analyzing the influence of platoon traveling speed, turning traffic flow and adjacent intersection distance to traffic signal coordinated control, an arterial group multi-objective decision-making model based on bi-level programming is established, where the up-level aims at maximum the area-wide total throughputs, and the low-level reduces the average vehicle delay. Structuring the restrained relation between offset, split and dynamic traffic volume, an arterial signal timing optimum proposal is given out, and the genetic algorithm is designed to optimize control strategy with the group multi-objective decision-making method is offered, which provides a new way of simultaneously accommodating to different traffic conditions.3. Neighboring intersection coordination rate, as a new traffic parameter, is put forward to quantitatively describe coordination between two neighboring intersections based on the theory of traffic flow continuity. The calculation formula of total area coordination rate is presented. By studying the intelligent coordinated control strategy of group of intersections, the framework of area-wide traffic signal optimal model is designed, and the coordinated rate based adjacent intersection optimizing control model is established, which combines with the traffic volume. Then, making the total area coordination rate acting as the judgment of signal timing, the multi-intersection group decision-making control optimizing model is presented, composing of common cycle, split and offset group utility functions. After that, the whole signal timing optimized process which is used to get the optimal scheme is proposed and a genetic-simulated annealing algorithm is design to optimize control strategy, which provides a new way and mentality to the design of area-wide traffic coordinated control.4. The dynamic hierarchical structure, physical architecture, and logical framework for area traffic signal control system are developed by contrast to all existing systems. Hardware components and function modules are described. The database, model library, knowledge base and method base of group decision support system are established, which determine the basis of choosing coordinated control model and optimal signal timing scheme, realizing the integration of group decision support system. The data stream and running process of the system is designed. Eventually, take traffic signal coordinated control systems on Ring Road, Nansha district, Guangzhou City, as an example, signal timing for each intersection, division process for control subareas, and the process to calculate optimal signal timing scheme are introduced. Scientific rationality of the division theory of control subareas and the method of coordinated control are verified by traffic simulation and practice.