Frankenstein’s PSO: An Engineered Composite Particle Swarm Optimization Algorithm

We introduce a high-performing composite particle swarm optimization (PSO) algorithm. In an analogy to the popular character of Mary Shelley’s famous novel, we call our algorithm Frankenstein’s PSO, as it consists of different algorithmic components drawn from other PSO variants. Frankenstein’s PSO constituents were selected after careful evaluation of their impact on speed and reliability. We present the process that guided us in selecting and adapting the algorithmic components included in the final version of the algorithm. The algorithm is validated through a comparison with other PSO variants on a number of well-known benchmark problems. Frankenstein’s PSO typically reaches high quality solutions faster and more frequently than the most commonly used PSO algorithms. We provide parameter selection guidelines for properly configuring Frankenstein’s PSO taking into account the requirements of the optimization task at hand.