Sptaio-temporal Co-location Congestion Pattern Mining And Application In Traffic Data

At present,most major cities of the world encounter the problem of traffic congestion,which is one of the“urban diseases”.Due to the continuous growth of per-capita car ownership,unreasonable road planning and inadequate traffic management,the problem of urban traffic congestion becomes more and more serious.Traffic congestion has not only causes huge economic losses,but also carries serious environmental pollution including the exhaust emission,noise pollution,etc.Therefore,it is essential to address the problem of urban traffic congestion.Although various research of solving the urban traffic congestion problem has been conducted,they do not consider the following two issues.First,the“traffic congestion”is a fuzzy state that cannot be simply determined by the unilateral factor,such as the speed of the vehicle,the capacity of road carrying or the length and width of the road.Second,there is a definite spatial correlation between urban roads in a certain period.So,we should consider the overall situation of urban roads rather than the congestion of someone certain road.Based on the above existing problems,this paper introduces the fuzzy set theory to define the level of congestion.Firstly,we obtain the congestion state of the road,then use the degree of membership in the fuzzy theory to measure the degree of road congestion,and propose a formula of calculating the membership value.Not only the problem of describing the state of road congestion,but also the quantitative analysis of the degree of congestion,can be addressed by our approach.At the same time,we introduce the time characteristic to the traditional spatial co-location pattern,and propose the concept of sptaio-temporal co-location congestion pattern.The most essential difference between the sptaio-temporal co-location congestion pattern and the traditional spatial co-location pattern is that both geospatial proximity and temporal proximity are considered at the same time.According to the above existing concepts,this paper firstly proposes the sptaio-temporal co-location congestion pattern mining framework and algorithm for discovering frequent congested patterns in urban roads.Then we design a traffic congestion visual prototype system and it can dynamically display the mining results.Finally,a feasible solution to solve the realistic congestion problem is proposed,it will have a great guiding role and practical significance.