CAJ | 학술논문

针对群搜索优化（GSO）算法存在的不足,提出一种新的GSO实现算法（NRGSO）.采用5个300维和7个30维的测试函数对NRGSO算法进行数值实验,并将其与GSO算法、微粒群优化（PSO）算法、遗传算法（GA）、进化规划（EP）、进化策略（ES）进行比较.结果表明,NRGSO算法的性能优于GSO算法;而在解决高维和多模态函数的优化问题方面,其性能优于PSO、GA、EP和ES等算法.NRGSO算法改进了群搜索优化原实现方法的不足,提高了算法的搜索性能,不仅在高维函数的优化中表现卓越,还能有效地避免陷入局部次优,并且在实际的优化问题中应用方便.
침대군수색우화（GSO）산법존재적불족,제출일충신적GSO실현산법（NRGSO）.채용5개300유화7개30유적측시함수대NRGSO산법진행수치실험,병장기여GSO산법、미립군우화（PSO）산법、유전산법（GA）、진화규화（EP）、진화책략（ES）진행비교.결과표명,NRGSO산법적성능우우GSO산법;이재해결고유화다모태함수적우화문제방면,기성능우우PSO、GA、EP화ES등산법.NRGSO산법개진료군수색우화원실현방법적불족,제고료산법적수색성능,불부재고유함수적우화중표현탁월,환능유효지피면함입국부차우,병차재실제적우화문제중응용방편.
A novel realization algorithm of group search optimizer （NRGSO） is proposed, aiming at overcoming the deficiency of GSO. And it is easier to be applied in practical problems. Five test functions of 300 dimensions and seven test functions of 30 dimensions are used to conduct the numerical experiments and the results of the novel algo-rithm are compared with those of GSO, particle swarm optimization （PSO）, genetic algorithm （GA）, evolutionary programming （EP） and evolutionary strategy （ES）. The algorithm proposed in this paper is better than GSO and its performance in solving the problems of high dimensions and multimodal functions is better than PSO, GA, EP and ES. NRGSO improves the original algorithm. It enhances its search ability and achieves better results. This novel algorithm performs excellently in functions of high dimensions, can effectively avoid being trapped in the local minima and is applicable in practical optimizer.