Using Improved Bidirectional A Algorithm to Achieve Optimal Path Planning

• Main Authors: Chin-HungLiu, 劉欽鴻
• Other Authors: Jer-Nan Juang
• Format: Others
• Language: en_US
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/37905576632426487083

碩士 === 國立成功大學 === 工程科學系碩博士班 === 100 === Over the last several decades, computer technology has drastically changed the landscape of the online world and the way we lead our life. Thanks to the advances in navigation technologies, it is now easy for people to travel to new places via path planning. However, at the same time, navigation devices may guide users to undesired places due to imperfect path planning algorithms. There are many path-planning methods we could use. We select A* algorithm to be our baseline approach and investigate how it arrives the optimal path in a known environment. A * algorithm is a kind of graphic search algorithms that merely depend on the logic operations. So its idea is to compute the cost of all nodes using a predesignated evaluation function, and the computed results are used to do optimal path planning. Its improved methods had also been studied in many papers. Since A* has a computational inefficiency problem, its improved algorithms had also been studied in some researches. In this thesis, we propose a new idea and implement all (improved) algorithms in order to compare their efficiency. We analyze the executing time, path, and the amounts of unknown expanded nodes. Taking the data structure into consideration,we replace the type people used with another data type to improve the search speed. After improving A* algorithm constantly, we propose our innovative method to get the shortest path more quickly. Finally, we implemented the simulation program with Java Program language and Android SDK, and then simulated it on the Android simulator.