摘要
针对麻雀搜索算法收敛速度慢、精确度不高、易陷入局部最优等问题,文中提出了一种融合多策略改进的麻雀搜索算法(improved sparrow search algorithm based on multiple strategies,ISSAMS)。通过引入Circle混沌映射初始化种群,增加种群的多样性,提高全局搜索能力;利用正余弦搜索策略,更新发现者的位置,选择最佳位置,增强局部搜索能力,避免陷入全局最优;加入萤火虫扰动更新最优个体位置,寻找可行解,提高局部搜索能力和寻优速度。为了验证算法改进的有效性,选取5个基准函数进行仿真实验,其中3个基准函数为单峰函数,2个为多峰函数,并与遗传算法、灰狼算法、粒子群算法和麻雀搜索算法比较。通过仿真实验结果表明基于多策略改进后的麻雀搜索算法具备跳出局部最优解能力,收敛速度更快,同时精确度更高,在对比其他4种算法后其总体性能更好。通过将该改进应用到优化BP神经网络的阈值和权值中,对比未改进的麻雀搜索算法优化的BP模型,其误差降低了14.73%,并且与另外3种算法优化的BP模型对比,其基于多策略改进的麻雀搜索算法优化模型的平均绝对百分比误差是最低的,效果最好,进一步验证了该改进算法的有效性。
Aimed at the problems such as slow convergence speed,low accuracy and easy to fall into local optimality,an improved sparrow search algorithm based on multiple strategies(ISSAMS)is proposed in this paper.Circle chaotic mapping was introduced to initialize the population,increase the diversity of the population,and improve the global search ability.Using sine and cosine search strategy was used to update the location of the finder,select the best location,enhance the local search ability,and avoid falling into the global optimization.Firefly disturbance was added to update the optimal individual position,search for feasible solutions,and improve the local search ability and search speed.In order to verify the effectiveness of the improved algorithm,five reference functions were selected for simulation experiments,among which three functions were single-peak functions and two functions were multi-peak functions,and compared with genetic algorithm,grey wolf algorithm,particle swarm optimization algorithm and sparrow search algorithm.The simulation results show that the improved sparrow search algorithm based on multiple strategies has the ability to jump out of the local optimal solution,the convergence speed is faster,the accuracy is higher,and the overall performance is better than the other four algorithms.By applying this improvement to the threshold and weight of the optimized BP neural network,the error of the BP model optimized by no improved sparrow search algorithm is reduced by 14.73%.Compared with the BP model optimized by the other three algorithms,the average absolute percentage error of the optimized model based on the multi-strategy improved sparrow search algorithm is the lowest and has the best effect.The effectiveness of the improved algorithm is further verified.
作者
薛涛
张安杰
XUE Tao;ZHANG Anjie(School of Computer Science,Xi’an Polytechnic University,Xi’an 710048,China)
出处
《西安工程大学学报》
CAS
2023年第2期96-104,共9页
Journal of Xi’an Polytechnic University
基金
陕西省技术创新引导专项计划资助项目(2020CGXNG-012)。