Hybrid particle swarm optimization algorithm for cost minimization in service-workflows with due dates
Yingchun Yuan · Journal of Communications · 2008
The cost minimization with due dates in service-workflow systems is an intractable problem. Traditional level-based algorithms restricted the due date within a time-window and the search scope was limited. HPSO (hybrid particle swarm optimization) algorithm was proposed by adapting a PSO algorithm effective for continuous problems. The proposed algorithm introduced operators of the particle randomly disturbing extremum, the position trace back and the crossover. By extending the research scope to the entire space, HPSO could obtain the optimal solution with a higher probability. Experimental results show that the HPSO reduces the cost 12.82% on average to that of the best currently available algorithm, as well as only 6.5s for a problem with 120 activities.