CAJ | 학술논문

针对入侵杂草优化算法（IWO）在迭代后期存在种群多样性差、局部搜索能力差、收敛精度不高等缺点，提出一种改进的入侵性杂草优化算法（IWODE）。首先，引入一个0~1之间的随机数，通过提高子代个体的质量来提高算法的寻优精度；其次，对繁殖后的新种群引入差分进化算法中的变异、交叉和选择策略以增加种群多样性，使算法在迭代过程中跳出局部最优，提高算法的全局寻优能力。对9个测试函数的仿真结果表明：无论是单峰还是多峰高维函数， IWODE 算法的收敛精度和稳定性都高于标准 IWO 算法。
침대입침잡초우화산법（IWO）재질대후기존재충군다양성차、국부수색능력차、수렴정도불고등결점，제출일충개진적입침성잡초우화산법（IWODE）。수선，인입일개0~1지간적수궤수，통과제고자대개체적질량래제고산법적심우정도；기차，대번식후적신충군인입차분진화산법중적변이、교차화선택책략이증가충군다양성，사산법재질대과정중도출국부최우，제고산법적전국심우능력。대9개측시함수적방진결과표명：무론시단봉환시다봉고유함수， IWODE 산법적수렴정도화은정성도고우표준 IWO 산법。
In view of the invasive weed optimization algorithm(IWO)in the diversity of the popula-tion difference iterative later,disadvantage of poor local search ability,convergence precision being not high,this study proposed one kind of improvement algorithm invasive weed(IWODE). In the al-gorithm,firstly,the introduction of a 0 ~ 1 between the random number improved the quality of the offspring,and the algorithm was used to improve the optimization accuracy;secondly,the post breed-ing new species introduced variation in evolutionary algorithms,and the crossover and selection strate-gy increased the diversity of population,which makes the algorithm jump out of local optimum in the iterative process,and improve the global searching ability of the algorithm;finally the simulation on 9 benchmark functions shows that whether it is single peak or multi peak high dimensional function,the convergence of IWODE algorithm accuracy and stability is better than the standard IWO algorithm.