Abstract:

According to the weakness of artificial bee colony algorithm(ABC), a new improved artificial bee colony algorithm(IABC) is presented and is applied to evaluate roundness error in minimum zone. The improved algorithm use information entropy to initialize population to enhance diversity, besides, a new search strategy is proposed in the stage of employed bees and onlookers. The fundamentals and implementation techniques of IABC are discussed. The optimal target function for roundness error evaluation and the fitness function of IABC are introduced. A series of classical test functions are selected in the experiments, the simulation results verifies the feasibility of IABC. Through several algorithms to measure some same sets of data for roundness error evaluation experiment, the results show that the evaluation precision of IABC is better than least square method(LSM)、genetic algorithm(GA)、particle swarm optimization(PSO) and some other algorithms, and it is superior to ABC in optimization of efficiency, quality and stability, the experiment results also show that IABC is correct and is a unified approach for roundness error evaluations.

Key words: error evaluation, minimum zone, roundness error, artificial bee colony algorithm(ABC)