Green Wave Coordinated Optimization Control Of Arterial Roads Based On Vehicle Trajectory Data

The arterial road is an important part of the urban road traffic system,carrying a large part of the urban traffic demand,especially the commuter demand,so its operation status determines the overall operation performance of the road network to some extent.At present,the main strategy to improve the traffic performance of arterial roads is to implement green wave coordination control.The specific strategies include real-time coordinated control and fixedtime coordinated control.The real-time green wave coordinated control strategy adjusts signal timing schemes in real time or dynamically based on the real-time acquisition of traffic state information.The fixed-time green wave coordinated control strategy establish targeted multiperiod signal timing schemes based on historical or measured empirical data.Both strategies have some limitations: the former requires high data quality and real-time accessibility,and the current technical methods are difficult to implement and expensive;the latter designs coordinated control scheme based on historical data or experienced data,which belongs to a relatively extensive strategy,and there are some problems such as the mismatch between the green wave range and the actual operation of the traffic flow as well as the insufficient utilization of the green wave.Considering the current means of acquiring traffic data,fixedtime green wave coordinated control optimization method for the urban arterial road based on vehicle trajectory data is studied.Aiming at the limitation that section test data cannot reflect traffic operation state spatially,an approach to estimate the cyclic traffic arrival flow rate is proposed based on the current mobile environment of vehicle trajectory data with low penetration rate.Firstly,arriving vehicles proportion distribution is calculated based on historical contemporaneous trajectory data.Then,two categories of vehicle arrival models are established to estimate the traffic arrival rate based on vehicle trajectory observations considering data conditions.Aiming at the situation where sufficient stopped trajectory data can be obtained,an estimation model of traffic arrival rate based on Poisson distribution is constructed,and the Expectation Maximization algorithm is used to solve it.Aiming at the situation where insufficient stopped trajectory data can be obtained,an estimation model of traffic arrival rate based on binomial distribution combining with penetration rate and traffic arrival information is constructed.Finally,a realworld intersection is selected for method verification,and results show that the mean absolute percentage error of the proposed method is 9.36%～12.64%.For further evaluation,the performance under different penetration rates and stopped trajectory ratios is analyzed.Results show that when the penetration rate is higher than 10%,the estimated result is relatively satisfactory with errors fluctuating around 10%,and when the stopped trajectory ratio is higher than 65%,the performance of the model based on Poisson distribution is better than that based on binomial distribution.Aiming at the shortcoming that traditional method cannot adjust signal timing parameters based on microscopic operation state of traffic flow,an green wave coordinated control optimization model is established based on traffic arrival information and traffic signal timing characteristics implied in vehicle trajectory data.Firstly,considering different trajectory sample morphology,a vehicle trajectory update model is established by capturing the exact mapping relationship between signal timing parameters,arrival flow rate and trajectory characteristic value such as vehicle arrival time based on shockwave theory.Then,with the goal of minimizing the total delay of sampled vehicles in arterial roads,an optimization model of bidirectional green wave coordinated control based on vehicle trajectory update is established.Finally,the Artificial Bee Colony algorithm is improved by introducing difference operator,which is used to solve the optimization problem accurately.The results of case study show that it is feasible to optimize the arterial coordination signal control scheme only with trajectory data,which can significantly reduce vehicle delay in coordination direction.