Abstract

ABSTRACT The booming development of the city and urban area made the transportation more and more complex. People even need a guidance to help them go from one place to another. The incoming huge needs of path finding can’t be only fulfilled by providing more computing resource and improving path finding algorithms yet. New methods for providing the routing solution are also needed to be considered carefully. This paper represents a concept to introduce Level-of-Detail technology for path finding process to reduce the algorithm’s computing resource consumption, and increase the algorithm’s efficiency. Keywords: Level of Detail, path finding, A* algorithm, region 1. INTRODUCTION The requirement of path guiding under restricted area such as underway and inner building is significantly increasing[1]. In the past, the researchers’ concentrati on is mainly on the evolution of the path finding algorithms. The development of the path finding algorithms can be classified into two main types: label correcting algorithms and label setting algorithms[2]. The label correcting algorithm is a routing algorithm which dynamically updates the basic moving area’s transporting value, while label setting algorithm does not change moving area’s transporting value after the label initialization of the experiment area. Until now label-correcting algorithm is still not the main stream of the current path finding researching, for the dynamic label updating costs huge computing resource during the path finding process, and the dynamic label changing significantly affect the real time responding of the path finding solution. Although the label-setting algorithm uses more computing resource on the initialization of the label matrix, the post path finding process doesn’t need to change the label values[2]. There are some famous classical label-setting algorithms, such as Dijkstra algorithm and A* algorithm. Dijkstra algorithm is an algorithm the shortest path is composed by one node in one step[3].The princi ple of Dijkstra algorithm is to find the shortest path in each iteration, and this principl e makes Dijkstra algorithm short sight of the movement in the whole area, for the shortest movement in one iteration may not be the right movement along the optimized path. And the A* algorithm somehow overcomes the shortness of Dijkstra algorithm[4]. A* algorithm not only intends to take the shortest moving action in each step, but also cares about the possibility of whether the expected step belonging to the best path from source to target. With the great advantage pr ovided by A* algorithm many present path finding solutions use A* algorithm as their core. The most famous optimized path solution using A* algorithm is bidirectional A* algorithm path finding solution. Bidirectional A* algorithm solution’s main idea is to use two A* algorithm searching process, which are the searching process from source to target and the searching process from target to source[5]. The two process operates independently at the same time, and the searching halts at once when the two searching process meet each other. This dual directional using of A* algorithm greatly reduces th e burden on memory cost caused by A* algorithm, and gets good result on the path finding efficiency[5]. Most current path finding algorithms introduced recently concentrate on the improvement of computing resource costs and computing time used for generating the optimized path. These algorithms can not meet the various people requirements in real world. There is a bad news that the current developing of path finding focuses too much on the optimization of algorithm, and ignores the real problem existed in the core part of path finding, which is the ignorance of analyzing people’s true demand of path guiding. And the real problem is that current path finding developing concentrates too much on what the path finding solution gives to the user, and doesn’t consider what the people really needs. For example, one user probably only needs a direction guide from position B to help him finish the organization

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call