Exploiting co-existence and co-evolution of mutualistic communities: A stable algorithm based on the plant-pollinator interactions

Abstract

In this paper, we have build-up a new optimization algorithm with novel stochastic search approach mimicking features of evolution and mutualistic cooperation observed among plants and animal pollinators in nature. The framework, based on the mathematics of the optimization mechanism of the plant-pollinator dynamics inspired the metaphorical design of the search strategies. We name our algorithm as a Plant pollinator optimization (PPO) and propose its variants by adding a few purposeful features of Evolutionary Algorithms. The competitive performance of the proposed approach and its variants is tested against 14 benchmark problems and superior performance of the proposed approach has been reported. The results are compared with that of hybrids of Particle Swarm Optimization adopted from literature. Subsequently, the results are used to draw statistical inferences regarding the convergence characteristics, exploratory features and consistency of the algorithm. The implementation of PPO on a realistic discrete optimization problem of dynamic scheduling and chartering of oil tankers further demonstrates the suitability of the proposed algorithm.