Real-Time Priority Control Of Tram Based On The Prediction Of Tram Dwell Time

Modern tram is widely considered as a popular and effective way of public transportation,it has the advantages of low construction cost,low construction difficulty,high safety coefficient,high environmental protection coefficient.Domestic trams generally adopt semi-independent right of way,the traditional traffic signal control strategy cannot guarantee the running efficiency of tram under the right of way.Tram signal priority can improve the running efficiency of tram,but it will have a negative impact on the social vehicles at the intersection,so how to design an effective signal priority control strategy for tram and reduce the impact on social vehicles while improving the running efficiency of tram will be a meaningful research topic.Based on the prediction of the dwell time of trams at stops,this article provides trams with the signal priority according to the real-time traffic flow at each intersection on the arterial streets.The Predictive real-time Tram Signal Priority(PRTSP)model based on dwell time prediction is established by analyzing the characteristics of tram real-time priority problem.The PRTSP model optimizes the total delay of trams and social vehicles in trunk lines and takes the effective green time of each phase and the offsets between intersections as the decision variables.The time for tram to reach the stop line at each intersection is one of the input parameters of PRTSP model,it is related to tram running time and tram dwell time at stops,the former is certain,while the latter is uncertain,it is non-linear related to the number of passengers getting on and off the bus and the headway with the former tram.In order to get more accurate tram dwell time at stops,this article uses the support vector regression machine to establish a predictive regression model for the dwell time of tram at stops.In order to solve the PRTSP model,an improved chaotic particle swarm optimization algorithm(PSO)is then designed.Chaos theory is introduced in the initial population stage and the mutation operator is introduced when the population optimal position is continuously stagnant,to improve the optimization ability.Finally,taking the L1 line of Wuhan optical valley tram as an example,a simulation experiment is designed.The experimental results show that the PRTSP model designed in this paper can effectively reduce the delay of social vehicles on the arterial streets while providing signal priority to tram.It also shows that the improved chaotic particle swarm optimization algorithm is better than the standard particle swarm optimization algorithm.