CAJ | 학술논문

针对粒子群优化算法易出现早熟收敛的问题，本文提出一种基于正交搜索的粒子群优化测试用例生成方法。首先，利用奇异值分解来预测种群的进化方向，在其正交方向进行搜索，可避免已搜索过的区域，有助于跳出局部最优；然后，对粒子速度项进行改进，使其与正交方向保持一致，保证种群可持续受到正交方向的影响，有利于减少奇异值分解次数，降低时间消耗；最后，对每代最优个体进行局部搜索，以增强算法局部搜索能力。实验证明，本文方法在覆盖率、运行时间、进化代数等指标上均有优势。
침대입자군우화산법역출현조숙수렴적문제，본문제출일충기우정교수색적입자군우화측시용례생성방법。수선，이용기이치분해래예측충군적진화방향，재기정교방향진행수색，가피면이수색과적구역，유조우도출국부최우；연후，대입자속도항진행개진，사기여정교방향보지일치，보증충군가지속수도정교방향적영향，유리우감소기이치분해차수，강저시간소모；최후，대매대최우개체진행국부수색，이증강산법국부수색능력。실험증명，본문방법재복개솔、운행시간、진화대수등지표상균유우세。
To solve the problem of premature convergence,this paper presents a method of generating test cases based on or-thogonal exploration and particle swarm optimization .First,singular value decomposition is used to estimate the evolution direction and drives the population towards orthogonal direction,so that our method can avoid searching those traversed areas so far and jump out of local optimum .Then,we change the velocity so that it is able to be consistent with the orthogonal direction,and as a result, the population can be affected continually,which can decrease the frequency of singular value decomposition and reduce the time consumption .Finally,the local search is used for the best particle in each generation .The experimental results show that our method has advantages in coverage,running time,and the number of generations .