CFO: A new compact swarm intelligent algorithm for global optimization and optimal bipedal robots walking

摘要

This paper introduces compact swarm intelligence. It presents a new estimation of distribution algorithm based on the swarm behavior of fireflies. The proposed algorithm is called the compact firefly optimizer (cFO). It uses a compact representation to store the population of the fireflies. Thus, the whole population is represented by a probabilistic distribution function. The compact swarm intelligence proposed allows the reduction of the computational capacity required to evolve the population toward the optimal solution. An important numerical test is used to evaluate the performance of the proposed algorithm. The results obtained show that the proposed compact firefly optimizer (cFO) is very competitive. Furthermore, the new concepts of compact swarm robotics are also introduced. The algorithm is applied to describe the first compact swarm robotics system. The objective is to realize an optimal stable walking by bipedal robot. Discussions, propositions and analysis let to think to a new important starting point for an emerging field of compact swarm intelligence, robotics, search and optimization.