An intelligent search strategy based on leadership, foraging efficiency and threshold response

Swarm intelligence (SI), inspired by collective behavior of animals, has drawn much attention for its advantages in solving complex engineering problems. There are currently various forms of algorithms derived from mimicking swarm behavior. A key component of such algorithms is the determination of how each individual in a population behaves. In this paper, the evolution of a population is determined by a novel decision-making model that combines the effect of Leadership, Optimal foraging and Threshold response (LOT) which is expressed as simple mathematical formulas. Based on these formulas, rules governing the decision-making process of each individual in a population are formulated. In order to verify the effectiveness of the proposed LOT decision-making model, five highly multi-modal functions of varying difficulties are tested together with the Particle Swarm Optimization (PSO) method. The LOT model has shown excellent performance in accuracy and global search ability compared with PSO.