摘要
针对随机产生输入权重和隐含层神经元阈值导致利用极限学习机辨识弹丸气动参数时会出现辨识结果发散问题,本文将粒子群算法与极限学习机结合,并且引入自适应更新策略以及粒子变异策略,提出了一种自适应变异粒子群优化极限学习机算法。该算法利用自适应变异粒子群算法寻优产生极限学习机的输入权重和隐含层阈值,有效改善算法性能。仿真实验表明,利用自适应变异粒子群优化极限学习机算法辨识弹丸气动参数,精度高、收敛速度快,能够充分满足实际工程需要。
In view of the identification results diverge when using extreme learning machine to identify the aerodynamic parameters of the projectile,due to the randomly generated input weights and hidden layer neuron thresholds,an adaptive mutation particle swarm optimization extreme learning machine algorithm is proposed.The paper introduces the adaptive update strategy and particle mutation strategy into particle swarm optimization algorithm and couples it with extreme learning machine.The proposed algorithm optimizes the input weights and hidden layer thresholds of extreme learning machine through adaptive mutation particle swarm optimization algorithm,the adaptive update strategy and particle mutation strategy in algorithm effectively improve the performance of the algorithm.Simulation experiments show that the use of adaptive mutation particle swarm optimization extreme learning machine exceedingly improve identification accuracy and convergence speed,which is practical in engineer application.
作者
夏悠然
管军
易文俊
XIA Youran;GUAN Jun;YI Wenjun(National Key Lab of Transient Physics,Nanjing University of Science and Technology,Nanjing 210094,China;School of Electronic and Information,Jiangsu University of Science and Technology,Zhenjiang 212100,China)
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2023年第2期521-529,共9页
Systems Engineering and Electronics
关键词
弹丸
气动参数辨识
极限学习机
粒子群优化算法
自适应更新策略
粒子变异策略
projectile
aerodynamic parameter identification
extreme learning machine
particle swarm optimization algorithm
adaptive update strategy
particle mutation strategy