Web Service Composition with Global Constraint Based on Discrete Particle Swarm Optimization

For semi-automatic composite services, operators make the universal services composite model that adapts itself to application demands according to given business background. This paper present independent global constrains Web service composition optimization methods based on Discrete particle swarm optimization (DPSO) and color Petri net (CPN). Firstly, a CPN modeling methods which can describe multi-attribute multi-constraint relations and associate relation between component services is presented. Secondly, taking advantage of the properties of CPN, DPSO is used to search a legal firing sequence in the CPN model, and the composite service corresponding to the legal firing sequence.Using legal firing sequences of the Petri net can make the service composition locating space based on DPSO shrink greatly, and avoid trapping in the local optimal solution. Theoretical analysis and experimental results indicate that this method owns both lower computation cost and higher success ratio of service composition.