Organizing tactics based optimization theory

Abstract

This paper proposed a new general framework for intelligent optimization based on organizing tactics rather than probability rules. Compared with the existing intelligent optimization algorithms, like Particle Swarm Optimization, this framework has several significant advantages. First, the “intelligence” does not depend on the probability rules of the operators, but their organizing tactics. Thus there are no probability equations that need to be updated, and involved control parameters are fewer, so it is easier to use in practice. Second, synergistic coexistence and automatic balance of the exploration and the exploitation are achieved in the running. Third, population diversity has been kept during the running. Fourth, most useless and ineffective repetitious operations are avoided, and thus the needed consumption of storage space and running time are lessened largely.