indexforcitytrajectories

indexforcitytrajectories内容摘要：

time. Moreover, explicitly storing a stream in its entirety is impossible, due to its unbounded nature. 数据流  数据流处理方法比传统的处理方法要快很多 .尽管该方法得到的结果并不精确 ,但是往往并不影响用户的最终决策 .  数据流的处理思路就是设计高效的单遍数据集扫描算法 ,在一个远小于数据规模的内存空间里不断更新一个代表数据集的结构 ——概要数据结构 ,从而实时、高效地获得近似查询结果 .  我将考虑采用一种比较简单的数据流处理方法来获得移动对象的聚集信息。 Type Definition  City trajectory work  Regions: regionid, MBR  roads: rid, ps,pe // start point,end point  lines: lid, lps,lpe // linear start point, linear end point.  CurrentAU: auid, ts,pos, length, v, m// start time, end time, the relative position of the polyline,v is a vector, the total moving object of the AU  PastAU: auid, ts,te, plid, pos, length, v, m  Mo: moid, (x,y), v // v is a vector currentAU operations  CurrentAU supports the following operations:  Create_AU() returns an AU at the entrance of a road segment.  the algorithm selects appropriate moving objects which have the same direction, nearly velocity and distance, to buildup an AU, and take the AU as one moving objects.  Drop_AU() is specific in AU index.  When vehicles goes out the road segment, their direction may change thus can not fit the roles of AU. So we drop it to disk for the AU records all details of the movements in this segment.  Delete_obj() and Insert_obj() is invoked when the position of the vehicle is outside of the AU at a certain time.  It finds the nearby AU in a greedy manner (including the original AU) to check yes or not it fit the rules for create AU. If an AU is fined, an Insertion will be operated, otherwise, a new AU is created for this moving object. Query  Window query  Range query  NN queries: ―find which object became the closest to a given point s during time interval T,‖  Aggregate queries: ―find how many objects passed through area Q during time interval T,‖ or, ―find the fastest object that will pass through area Q in the next 5 minutes from now‖  similarity queries: ―find objects that moved similarly to the movement of a given object o over an interval T‖ Related wor。