Abstract :
Particle swarm optimization (PSO) simulates the boids´ collective behaviors. The original biological background of boid should follow three basic simple steering behaviors: separation, alignment and cohesion. However, to promote a fast convergent speed, the velocity update manner of each boid omits the alignment rule, this may result premature convergence phenomenon. Therefore, in this paper, the alignment rule is added to the velocity update manner for optimizing the multi-modal numerical problems, in which each particle adjusts its moving direction according to the personal historical best position and the alignment direction. Furthermore, a mutation operator is also introduced to enhance the population diversity. Simulation results show the proposed algorithm is effective and efficient.
Keywords :
particle swarm optimisation; alignment particle swarm optimization; multimodal numerical problem; mutation operator; population diversity; simulation result; Algorithm design and analysis; Biological system modeling; Convergence; Cultural differences; Gaussian distribution; Genetic mutations; Particle swarm optimization; Random number generation; Size control; Velocity control; alignment rule; multi-modal numerical problems; mutation operator;
