Research Of Iterative Learning Control Based On Unfalsified Strategy For Urban Arterial Roads

With the rapid progress of urbanization in China,traffic congestion has become one of the inevitable and nonnegligible problems in urban management.The arterial roads normally bear the main traffic pressure of the urban transprotation system.Therefore,to heighten the coordinated control effect of urban arterial roads and and to reduce the traffic delay and parking rate on arterial roads are of great significance for improving the traffic condition of the whole city.The existing arterial road signal control methods play the vital roles in field application.However,there are still some problems.For example,the computational complexity increases exponentially with the expansion of the network scale,and the optimal control effect depends heavily on the precision of the mathematical model,but the exact model is difficult to obtain for the complex arterial road road traffic system.To partially overcome the aforementioned problems,the iterative learning control method is introduced into the urban arterial road control.An iterative learning control method based on unfalsified strategy for Arterial Roads is proposed.The simulation results show that these algorithms have strong self-adaptive and self-learning habits,and can solve urban traffic congestion more effectively than the traditional methods.The main contents of the thesis are summarized as follows.1.With the modification of the store-and-forward traffic model,a coordinated control method is proposed for single intersections and the arterial roads.To be specific,by considering the traffic streams as the elementary components of the traffic,the traffic model is refined first from road segments into relevant traffic streams to improve the modeling.Then the iterative learning control method is applied to achieve traffic signal control for both single intersection and the arterial road.Also,by considering the interaction between intersections as the measurable disturbance,the coordination of intersections along the arterial road is conduceted in a pairwise manner.2.The unfalsified strategy is introduced into the design procedure so as to adaptively tune the closed-loop learing rate of ILC-based control law according to the real-time change of the traffic demand.In turn,the robustness the quick responsibility of the system can be improved.Moreover,the convergence of the system is further analyzed in cases of the varations of external disturbances and the mismatch of the repetive condition of the initial state for the ILC-based design.3.The aforementioned methods are testified by using matlab and vissim with the actual network setting and real filed data of Keqiao District,Shaoxing.The effectiveness of the methods are illustrated.Finally,the research contents of the thesis are summarized,and the perspectives of future works are proposed.