PROJECT TITLE :

An Efficient Ride-Sharing Framework for Maximizing Shared Route - 2018

ABSTRACT:

Ride-sharing (RS) has nice values in saving energy and assuaging traffic pressure. Existing studies will be improved for higher efficiency. Therefore, we propose a new ride-sharing model, where each driver incorporates a requirement that if the driver shares a ride with a rider, the shared route percentage (i.e., the ratio of the shared route's distance to the motive force's total travel distance) exceeds an expectation rate of the driver, e.g., zero.8. We contemplate 2 variants of this downside. The first considers multiple drivers and multiple riders and aims to compute driver-rider pairs to maximize the overall shared route percentage (SRP). We have a tendency to model this downside as the most weighted bigraph matching problem, where the vertices are drivers and riders, edges are driver-rider pairs, and edge weights are driver-rider's SRP. However, it's rather expensive to compute the SRP values for massive numbers of driver-rider pairs on road networks. To handle this drawback, we tend to propose an efficient technique to prune many unnecessary driver-rider pairs and avoid computing the SRP values for each pair. To enhance the efficiency, we propose an approximate method with error certain guarantee. The basic idea is that we have a tendency to compute an upper certain and a lower certain for each driver-rider pair in constant time. Then, we estimate an upper bound and a lower bound of the graph matching. Next, we tend to choose some driver-rider pairs, compute their real shortest-route distance, and update the lower and upper bounds of the maximum graph matching. We have a tendency to repeat higher than steps till the ratio of the higher bound to the lower certain is not larger than a given approximate rate. The second considers multiple drivers and one rider and aims to seek out the prime- k drivers for the rider with the most important SRP. We tend to initial prune a massive number of drivers that can't meet the SRP needs. Then, we tend to propose a best-first algorithm that progressively selects the drivers with high probability to be within the high- k results and prunes the drivers that cannot be within the top- k results. Intensive