Dynamic shortest path with probability

Abstract

Shortest path algorithm is necessary when people get lost or do not know how to reach their destinations. They can rely on the shortest path application itself and search for the best route to guide them the way to reach the destination. In case people are in hurry and they have no idea of how to reach the destination, then shortest path is the useful algorithm to help them solving this problem. Unfortunately, the classical shortest path algorithm will only calculate the shortest path base on the reference weight value. So within the rush hours, the shortest path data may not be accurate enough. Imagine that within the rush hours, if people are wake up late, they may miss their regular bus to their office. In this case, they do not have any idea how to reach to their offices on time as usual. So should they wait for the next bus or change to the other transportation? In this question people want to know two answers for this special case: (1) What is the best route to reach to the office? (2) What is the probability of arriving in accurate when people change to other transportation? Hence, if there has a system that can compute the shortest path base on the statistic information of the traffic during the rush hours, and provide the suggested route with the probability to reach the office within the time constraint will definitely help people to solve the problem. This project is aims to develop the algorithm to compute the shortest path under the dynamic network environment. Moreover, the shortest path computed by the algorithm must fulfill the time constraint which specified by the users.