Variable Speed Limit Control Technique On Expressways

With the rapid increase of motorized traffic demand, expressways have suffered from severe traffic congestion and safety problems. A reasonable speed limit could help keep driver speeds within the safe level under current road traffic and environment situations. However, traditional determining method cannot provide safe and reasonable speed limits for drivers under varying traffic situations. Thus naturally it belongs to a static speed limit method. The variable speed limit (VSL) control technique detects automatically the traffic flow parameters on expressways and estimates the traffic states based on real-time data. The VSL adjusts automatically the speed limits according to some control strategies to intervene the operation of traffic flow, in order to achieve the improvement of traffic efficiency and safety. Thus the VSL is a dynamic speed limit method.As a novel expressway mainline traffic control technique, the VSL has recently drawn a great interest from researchers and has achieved some effects in practical engineering applications. However, only a few studies have focused on the critical theoretical issues in the VSL control. The theoretical foundation of VSL has not been well established, which restricts the development of the novel technique. This study can help better understand how VSL control affects traffic flow and how various optimal control algorithms affect the effectiveness of VSL for improving expressway safety and efficiency. For practical applications, the study can help improve the level of expressway traffic management and reduce the severe problems on expressways.The study tries to explore and analyze the critical issues associated with the expressway VSL control technique. The work was sponsored by the National Natural Science Foundation of China (No.51322810). The study collected data from Europe, America and China expressways, and analyzed the study tasks from the mechanism, technique, evaluation, and example aspects. The study first analyzed the impacts of VSL on the traffic operation, and developed the traffic simulation technique with the VSL control. Then the study proposed several VSL strategies and optimizing algorithms, and evaluated the effectiveness of VSL on improving expressway safety and efficiency. At last, the study validated the control effects of VSL strategies under the particular environmental and traffic flow conditions on expressways in China. A method for determining the optimal deployment location of VSL signs was also developed. More specifically, this study includes the following contents:First, the study analyzes the impacts of VSL control on traffic operation based on actual data collected on expressways. The characteristics of time headway, vehicle speed and driver compliance to speed limit are analyzed. The impacts of VSL on key traffic flow parameters and fundamental diagram are evaluated. The mechanism of improvement of traffic efficiency under VSL control is analyzed. The study also discusses the relationship between expressway traffic flow and traffic safety, in order to analyze the mechanism of improvement of traffic safety with the VSL control.Second, the study develops the traffic simulation model for the VSL studies. The critical consideration of traffic simulation with VSL control is discussed. Then the cell transmission model (CTM) is modified and improved to simulate the bottleneck capacity drop phenomenon, the stop-and-go traffic wave in congestion, and the impacts of VSL on traffic flow with different driver compliance rates. The calibration and validation method for the modified CTM with VSL control are proposed, and examples are given.Third, the VSL optimal control problem is divided into the single objective optimization and the multi-objective optimization. The genetic algorithm based optimization of VSL control strategy is proposed to explore the optimal solution during the evolution process. The reinforcement learning based optimal VSL control strategy is proposed. The agent learns the optimal speed limits under various traffic states through the training process. The multi-objective optimization theory is integrated with the multi-objective VSL control to explore the Pareto solutions using the multi-objective genetic algorithm.Fourth, the optimal control of VSL for improving expressway safety is proposed. Two road facilities including the recurrent isolated bottleneck section and the continuous bottleneck section are analyzed. Real-time crash risk prediction models are proposed for different facilities and the VSL control strategies are proposed. Then the VSL control effects are optimized with the genetic algorithm and the reinforcement learning algorithm separately. The control effects between different optimizing algorithms are also discussed.Fifth, the optimal control of VSL for improving expressway efficiency is proposed. The study first develops a modified feedback based VSL algorithm and evaluates its control effects for a merge isolated bottleneck section. Then the optimal control effects of VSL based on the reinforcement learning algorithm are evaluated. The mechanism and effects of different algorithms are compared. The study compares the control effects of VSL and ramp metering algorithms at different types of bottleneck sections. A coordinated algorithm is proposed to achieve the optimal control effects. A method is developed to help select optimal algorithms for different bottleneck sections under various traffic flow conditions.Sixth, the multi-objective optimal VSL control problem is studied. The VSL strategy that aims at improving both expressway safety and efficiency is proposed. The multi-objective genetic algorithm is conducted to determine the Pareto solution for the VSL optimal control problem. The features of Pareto front under different deployments of VSL signs and driver compliances are analyzed, to provide decision support for the selection of optimal strategies.Seventh, the effectiveness of VSL control strategies is evaluated for expressways in China. The road geometric characteristics, traffic operational features and mechanism of congestion occurrence on Chinese expressways are analyzed. Then the VSL control systems are designed accordingly for the selected bottleneck sections. Based on the calibrated simulation model, the effectiveness of VSL control strategies on reducing travel time and crash risk on Chinese expressways are evaluated.Finally, the method for determining the optimal deployment locations of VSL signs on expressway mainline is proposed. The hazardous bottleneck section is identified according to the crash risk model. Several scenarios with different lengths of VSL controlled sections and densities of VSL signs are proposed. The VSL control effects in each scenario are optimized. The cost-benefit analyses are conducted to help determine the optimal deployments of VSL signs on expressway mainlines.