Memetic algorithm for the resource-constrained project scheduling problem

In this paper we propose a memetic algorithm (MA) for the well-known resource-constrained project scheduling problem. The MA employs the activity list representation, makes use of one-point crossover as recombination operator and five local search procedures to improve each off-spring individual. A detailed computational experiment is set up using standard instances from the problem instance library PSPLIB. Computational results show that, using 50 000 generated schedules as a stop condition, the proposed MA outperforms all the compared state-of-the-art heuristics for the instance data set J60 and that it is competitive with the compared heuristics for the instance data sets J30 and J120.