Impact of Acceleration Coefficient Strategies with Random Neighborhood Topology in Particle Swarm Optimization

Impact of Acceleration Coefficient Strategies with Random Neighborhood Topology in Particle Swarm Optimization Mrs. Snehal Mohan Kamalapur, Dr. Varsha Hemant Patil Research Scholar, Research Guide DYPIET, Vice Principal, MCERC, Pune , Nashik India Abstract: Particle Swarm Optimization is optimization technique having few parameters to tune. Inertia Weight with velocity clamping has great impact on the PSO performance. Velocity clamping is simple and effective solution to swarm explosion while inertia weight has an effect on swarm diversity. Another PSO performance enhancement parameter is the neighborhood topology. Stochastic acceleration terms c1 and c2 pulls each particle towards personal best and global best positions. The linearly decreasing scheme of inertia weight and velocity clamping are considered herein the paper for swarm diversity and explosion respectively. The work under consideration extends the existing parameter setting techniques on inertia weight and acceleration coefficients. The concept of a "dynamic" neighborhood is explored. The paper focuses on linear varying inertia weight, random acceleration coefficient and random neighborhood topology link based variants. Experiments have been conducted to evaluate the performance.