Research On The Detection And The Intervention Method Of Taxi Speeding Behavior Based On GPS Trajectory Data

Speeding,as a typical traffic violation,seriously affects road traffic safety.Among them,taxis are an important part of urban road traffic participants,and their speeding behavior is also very serious.As taxi drivers are very familiar with the urban road network and work intensively,in order to achieve more trip orders in a limited time to increase revenue,this driving group has a huge speeding demand.However,the current penalties and interventions for taxi speeding are mainly carried out from speeding enforcement and the management of rewards and punishments within taxi companies.The former treats taxi as social vehicle to carry out speeding enforcement,but this method is difficult to supervise the taxi speeding behavior throughout the entire trip,and the probability of detection of taxi speeding behavior is low;the latter uses a single performance scoring mechanism to punish taxi drivers,the penalty standard is single.Therefore,the effects of the two methods above of speeding intervention are limited.Therefore,it is necessary to use new techniques to improve the probability of detection and punishment of taxi speeding behavior,and to construct a new speeding behavior punishment mechanism based on the characteristics of taxi drivers’ speeding behavior,and to fully intervene in taxi drivers’ speeding behavior.Therefore,this article used vehicle-mounted GPS technology to obtain vehicle GPS trajectory data by designing a speeding behavior detection driving test plan,and built a speeding behavior detection algorithm based on continuous speeding GPS trajectory points and based on the average speed between two points,the detection accuracy and advantages and disadvantages of the two algorithms were analyzed and compared.Considering the construction of a hybrid speeding behavior detection algorithm combining the two algorithms above,the optimal algorithm was confirmed.Based on the optimal detection algorithm and the taxi GPS trajectory data of a certain taxi company in Chengdu in the first week of November 2016,the taxi speeding behavior was extracted,and the distribution characteristics of taxi speeding behavior were analyzed from the number and range of speeding.On this basis,according to the step pricing theory,the taxi speeding behavior step-decreasing performance scoring mechanism was designed,and the economic price elasticity concept was introduced to analyze the response of taxi drivers to the step punishment mechanism,and the taxi speeding behavior step-decreasing performance scoring model was constructed,using taxi speeding data in Chengdu as an example for application was analyzed based on the model.Taking into account the delay and discontinuity of the intervention effect of the methods above over speeding behavior,the vehicle-mounted notification system was used to study the speeding intervention method based on dynamic prompt information.The research results showed that the accuracy of the detection algorithm based on the trajectory point sequence composed of the average speed between 3 adjacent GPS trajectory points and at least two consecutive speeding trajectory points as the condition for judging the occurrence of the speeding behavior is 88.46%,which was high in accuracy and met the detection requirements demand.Taxi speeding behavior was mainly concentrated in the low and medium speeding range and was frequent occurred.The step-decreasing performance scoring mechanism had a strong deterrent effect on drivers with different numbers and ranges of speeding.The greater the number of speeding and the greater the speeding range were,the greater the penalty intensity was,which could comprehensively suppress the speeding behavior of taxis.The real-time speeding intervention method based on voice and text informing the content of speeding penalties was proved to be the best intervention.