Improving electromagnetism algorithm for solving resource allocation problem in stochastic networks

This study investigates optimal resource allocation for minimizing total cost in stochastic networks (SN) where the duration of all activities involved is not only a random variable, but also a function of the resources allocated. The total cost of the network comprises resource usage cost and penalty cost. An Electromagnetism Algorithm (EA) is used as a decision tool for optimization and a Label-Correcting Tracing Algorithm (LCTA) for approximation of completion time in SN is suggested. Furthermore, the Critical Path Cluster Algorithm (CPCA) and Cluster Local Search Algorithm (CLSA) are developed to enhance EA's search ability for resource allocation. Results from numerical experiments show that the proposed EA yields good solution quality.