Catalyze Sharing Economy: Optimized Multi-Task Allocation for Urban Transport Crowdsourcing

摘要

Urban transport crowdsourcing is a new paradigm where a crowd of users share vehicles to carry other people or packages when they are on their way to someplace. Different from the traditional location based crowdsourcing system (e.g. mobile crowdsensing, etc.), the task has to be finished with passing two different locations (i.e. start and end points), so general task allocation algorithms cannot be applied to urban transport crowdsourcing directly. In this paper, we focus on the multi-task allocation in urban transport crowdsourciang with a more realistic scenario. We propose a heuristic greedy algorithm called Saving Most First (SMF) which is simple and effective to assign tasks. Then, to jump out of the local optimal result, an optimized SMF based genetic algorithm (SMF-GA) is devised. Finally, we demonstrate the performance of SMF and SMF-GA with extensive simulations based on a large scale real vehicle traces. The result indicates that SMF-GA can save longer distance, converge more quickly and motivate more participants than other contrast algorithms.