Research On Smart Generation Method Of Multi-Period Two-Way Bus Green Wave Scheme For Urban Arterial

Conventional public transit plays a vital role of urban public transport,and it has the advantages of large passenger capacity,low cost,low energy consumption,low pollution and less resource occupation per capita.In the process of urban road traffic management and service,reducing the number of bus stops at urban road intersections through bus traffic signal priority technology is of great significance to improving the level of urban road network bus services.The bus green wave control of arterial is an important means of bus signal priority,and it is also one of the easier methods to realize in the existing traffic signal control system.Aiming at the constraints existing in the current bus green wave optimization model,which do not consider the uncertainty of bus travel time and the modeling of the initial queuing and dissipation time of red light vehicles in the coordinated phase of downstream intersections,this paper estimates the uncertainty distribution of bus travel time in multiple time periods.Based on the estimation of the uncertainty distribution of bus travel time in multiple periods,combined with the vehicle queuing dissipation time model at the downstream intersection in the red light phase,the traditional bus green wave optimization model MAXBAND is improved,and the performance of the model is compared and evaluated.The specific research work of the thesis includes:First,considering that the travel time of bus vehicles from the upstream intersection to the downstream intersection is affected by the demand of passengers getting on and off the bus and the fluctuation of the traffic flow operation state,this paper proposed a Hidden Markov-based bus travel time uncertainty estimation model.Considering the operation characteristics of bus vehicles on the urban road,the road section is discretized to form several road section homogeneous cells;the bus vehicle operation state is set to two hidden states of normal driving and platform parking,based on the hidden Markov model and the relationship between the upstream and downstream of the road section,the probability model of the joint distribution of the travel time of bus vehicles at the cellular level is constructed.Taking the travel time series of multiple cells in the time period as the input,a Baum-Welch(BW)algorithm-based method was proposed to the parameter learning of the probability model of the joint distribution of bus travel time.Monte Carlo sampling is performed on the joint distribution of bus travel time after parameter calibration,and the section-level bus travel time samples are obtained through aggregation,and the uncertainty of bus travel time is quantified in the form of probability distribution.The results of case analysis based on real data show that the uncertainty estimation model of bus travel time proposed in this paper can better consider the operation characteristics of bus vehicles,and can quantitatively estimate the travel time and uncertainty distribution of bus vehicles more accurately.Secondly,for the urban arterial road where social vehicles and public vehicles are mixed,with the goal of reducing the number of stops and delay time of public vehicles at intersections,the paper considers the uncertainty of the travel time of public vehicles,and proposes a method based on the classic MAXBAND method.Bus green wave optimization model.Considering the possible different scenarios of the starting time of the coordinated phase traffic lights at the upstream and downstream intersections,a specific calculation formula for the initial queuing length of social vehicles during the coordinated phase red lights at the downstream intersections is proposed,and a segmented model of the initial queuing dissipation time of social vehicles in the scenario of coordinated phase difference between upstream and downstream intersections is constructed;based on multi-period bus travel time uncertainty probability distribution and social vehicle initial queuing dissipation time model,the classic MAXBAND model is improved,and a two-way bus green wave optimization considering uncertainty variables is constructed.Aiming at the characteristics of the intersection queuing dissipation time piecewise function,the auxiliary variables were used to linearly transform the constructed MAXBAND constraints,and a MAXBAND optimization solution method combining Monte Carlo and genetic algorithms is proposed.Finally,taking the Qianjin Road in Kunshan City as the object,based on the real data,the paper evaluated and analyzed the green wave optimization scheme of public transit in different scenarios(peak and off-peak)based on the SUMO traffic simulation software.The evaluation results show that the improved MAXBAND model proposed in this paper considering the uncertainty of the travel time of public transit vehicles can effectively reduce the average delay,the number of stops and stopping-and-passing ratio.During peak hours,vehicle delays are reduced by 24.6%,the number of stops and the stopping-and-passing ratio are reduced by 50% and 50%,respectively.During the offpeak hours,the average delay of bus vehicles is reduced by 38.8%,and the number of stops and the stopping-and-passing ratio are reduced by 24% and 25% respectively.Compared with the consideration of time average,the operation of public transit vehicles is more stable and less volatile.The standard deviation of vehicle delays at peak and offpeak hours decreased by 61% and 41.3%,respectively,and the number of stops and the stopping-and-passing ratio were not significantly different.