Parallel Differential Evolution with Variable Population Size for Global Optimization

The results of evolutionary algorithms depends on population diversity that normally decreases by increasing the selection pressure from generation to generation. Usually, this can lead evolution process to get stuck in local optima. The study is focused on mechanisms to avoid this undesired phenomenon by introducing parallel differential evolution that decompose a monolithic population into more variable-sized sub-populations, which evolve independently of each other. The proposed parallel algorithm operates with individuals having some characteristics of agents, e.g., they act autonomously by selecting actions, with which they affect the state of environment. This incorporates two additional mechanisms: aging, and adaptive population growth, which direct the decision-making by individuals. The proposed parallel differential evolution was applied to the CEC’18 benchmark function suite, while the produced results were compared with some traditional stochastic nature-inspired population-based and state-of-the-art algorithms.