CAJ | 학술논문

针对现有粒子群算法缺乏优化问题的先验信息,粒子搜索具有盲目性的问题,提出了一种基于邻域粗糙集模型的改进粒子群优化算法。该算法利用邻域粗糙集的方法获得粒子群寻优空间的先验信息,并动态地缩小粒子群的寻优区域,达到了更高效率地进行粒子群寻优的目的。通过优化10个典型测试函数,并与LDW-PSO优化过程进行比较验证了算法的有效性。
침대현유입자군산법결핍우화문제적선험신식,입자수색구유맹목성적문제,제출료일충기우린역조조집모형적개진입자군우화산법。해산법이용린역조조집적방법획득입자군심우공간적선험신식,병동태지축소입자군적심우구역,체도료경고효솔지진행입자군심우적목적。통과우화10개전형측시함수,병여LDW-PSO우화과정진행비교험증료산법적유효성。
In view of the disadvantages of the particle swarm optimization, such as the prior information deficiency ofoptimization problems and the blindness of particle searching performance, an improved particle swarm optimization algorithm based on the neighborhood model and rough set theory was proposed. Based on the neighborhood method, the prior information of the particles＇ searching space was obtained, the optimization region of the swarm dynamically narrowed and the efficiency of the optimization was improved. Comparison studies were done for the LDW-PSO and the proposed method. The experimental results for ten classical test functions prove the effectiveness of the improved algorithm.