ABSTRACT
This chapter is devoted to the theoretical foundation of particle swarm optimization, the latest PSO algorithm advances, neural network optimization methods and illustrative examples. The standard bio-inspired particle swarm algorithm is presented and some important variations of the PSO algorithm are explained and analyzed such as: Quantum PSO algorithm with delta potential well and harmonic oscillator, Chaotic PSO, mutation-based PSO and multi-objective PSO. Advanced aspects and variants of particle swarm optimization are discussed, such as: swarm topologies, boundary approaches, swarm initialization techniques, stopping criteria and swarm mutation operators, bringing to the reader a complete presentation of the PSO algorithm and its latest variations. Illustrative application examples are presented regarding unimodal, multimodal, convex and non-convex function approximation.
