Transportation Cloud Service Composition Based on Fuzzy Programming and Genetic Algorithm

Abstract

The push toward smarter transportation management and decision-making has increased significantly in recent years. Toward this end, it is constructive to establish a traffic intelligence platform leveraging web services, big data analytics, and cloud computing. The developers of any such platform always face the challenge of selecting or creating composition plans from among numerous possible plans, often with unclear requirements, that satisfy their quality-of-service (QoS) requirements. Typical QoS properties associated with a web service are execution cost and time, availability, successful evaluation, usage frequency, and accuracy. Most of these factors are vague and/or difficult to quantity. To address the need to handle vague inputs, a constraint satisfaction-based web service composition algorithm is proposed, which is based on fuzzy multi-objective linear programming. A genetic algorithm with Pareto optimization evaluation with weighted standardized Euclidean distance is used that demonstrates good performance as the quantity of service candidate increases. The algorithm is capable of finding a satisfactory solution under input of vague QoS requirements, which shows better capability than multiple-objective linear programming. The applicability and performance of the proposed algorithm are validated in an online transportation analytics platform environment.