Coarse-Grained Path Planning Under Dynamic Situational Environment

Abstract

With the enrichment of data acquisition means, large amounts of spatiotemporal data in the traffic network are accumulated in real-time by various sensors and multimedia devices. Through status identification, geographic mapping and statistical calculation, a dynamic situational road network is formed. Existing path planning methods based on the static network are inadequate in coping with the traffic situation, which seriously affects the performance of the path planning in accuracy and ETA. Considering the shortage of above, this paper proposes a coarse-grained path planning method under dynamic situational network. By introducing streaming graph partition and constructing a hierarchical index, the computational cost of path planning is effectively reduced. The pathfinding scope is decreased according to the coarse-grained path by progressively shrinking the minimum containing subarea, converting the pathfinding to a small-scale addressing in the hierarchical index. Meanwhile, the situation update and link reconstruction are performed within each tracking subarea, thus making dynamic perception of the traffic situation during travel. Finally, we conduct simulation experiments on the Beijing traffic dataset to verify the effectiveness of the proposed method.