Research On Time-Dependent Popular Route Recommendation

In the last few years,the rapid development of national economics increases the popularity rate of automobiles.The use of cars expands the scope of human activity and is convenient for people to travel.However,it brings new difficulties and challenges.The rapid growth of vehicle number,and the lagging of road resources construction lead to more and more serious traffic congestion and consume a lot of opportunity cost.One of the effective measures to solve the traffic problems is to design reasonable travel routes.The reasonable route,not only can make people get a better driving experience,but also can effectively avoid the occurrence of congestion.Thereby,the purpose of saving time and protecting the environment can be achieved.Although the research on route recommendation has existed for a long time,most of the related work remains in the static network.In order to solve the above problem,this thesis proposes a new time-dependent model of popular route recommendation.The obvious difference between the model and the existing models is that it takes into account the time-dependent feature of road network.Besides,it mines the real effective driving experience from taxi history data and recommends the experience to the general user.The main contributions of the thesis include:First,the thesis proposes a specific algorithm to set up the time-dependent function of road against the time-dependent characteristics of dynamic network.Firstly,it divides GPS dataset into different parts according to time frames and mines the traffic flow transition probability distribution within each time frame.Secondly,it clusters the time frames and gets the time-dependent popularity function.Second,aiming to the problem existing in traditional road searching method,which is the unification of accuracy and efficiency,the thesis puts forward a new method to construct local road network.In the course of it,the conception of out-in degree trajectory linked list is proposed and the local road network is constructed based on this conception.Meanwhile,in order to improve the efficiency of building a local road network,two types of pruning methods are introduced,which are the dominant trajectory pruning method and angle offset interval pruning method respactively.Finally,for the purpose of higher route popularity and shorter driving time,the thesis proposes three kinds of the most popular route recommendation methods,which are the most popular route recommendation method based on improved genetic algorithm,the local optimum popular route recommendation method and the global optimum popular route recommendation method respectively.The experimental results validate the effectiveness and the practicability of the time-dependent popular route recommendation method.