摘要
针对麻雀搜索算法在迭代后期种群多样性减少、易陷入局部最优等问题,提出改进麻雀搜索算法(ISSA)。引入Sobol序列,提高初始种群的多样性;引入黄金正弦算法,平衡全局搜索和局部开发能力;引入高斯差分变异,提高种群跳出局部最优的能力。10种基准函数的测试结果表明,ISSA有着更好的寻优精度与收敛速度。使用ISSA对SVM的超参数进行寻优,构建分类模型并应用于断路器故障诊断,验证了该方法在工程应用上的可行性。
Aiming at the problems that the population diversity of sparrow search algorithm decreases and it is easy to fall into local optimization in the late iteration,an improved sparrow search algorithm(ISSA)was proposed.The diversity of initial solution was improved by introducing Sobol sequence.The abilities of global search and local development were balanced by intro-ducing the Golden sine algorithm.The ability of the solution to jump out of the local optimum was improved by introducing the Gaussian difference mutation.The test results of ten benchmark functions show that ISSA has better search accuracy and convergence speed.The hyperparameters of SVM were optimized using ISSA,the classification model was constructed and applied to circuit breaker fault diagnosis.The feasibility of this method is verified in engineering application.
作者
石颉
杜国庆
SHI Jie;DU Guo-qing(School of Electronic and Information Engineering,Suzhou University of Science and Technology,Suzhou 215000,China)
出处
《计算机工程与设计》
北大核心
2023年第3期954-960,F0003,共8页
Computer Engineering and Design
关键词
改进麻雀搜索算法
Sobol序列
黄金正弦算法
高斯差分变异
支持向量机
参数优化
故障诊断
工程应用
improved sparrow search algorithm
Sobol sequence
golden sine algorithm
Gaussian difference mutation
support vector machine
parameter optimization
fault diagnosis
engineering application
