A quasi-oppositional learning of updating quantum state and Q-learning based on the dung beetle algorithm for global optimization

Abstract

There are many tricky optimization problems in real life, and metaheuristic algorithms are the most effective way to solve optimization problems at a lower cost. The dung beetle optimization algorithm (DBO) is a more innovative algorithm proposed in 2022, which is affected by the action of dung beetles such as ball rolling, foraging, and reproduction. Therefore, A dung beetle optimization algorithm is proposed based on quasi-oppositional learning and Q-learning (QOLDBO). First, the quantum state update idea is cleverly integrated into quasi-oppositional learning to increase the randomness of the generated population. And the best behavior pattern is selected by adding Q-learning in the rolling stage to improve the search effect. In addition, the variable spiral local domain method is proposed to make up for the shortage of developing only around the neighborhood optimum. For the optimal solution of each iteration, the dimensional adaptive Gaussian variation is selected and the optimal solution is retained. Experimental performance tests show that QOLDBO performs well in both benchmark test functions and CEC 2017. Simultaneously, the validity of the algorithm is verified on several classical practical application engineering problems.