摘要
针对天波雷达方位分辨力低和传统解析算法定位误差较大的缺点,提出一种混沌变异灰狼算法优化核极限学习机(KELM)的定位模型。首先,该模型将分段线性混沌映射、自适应柯西变异和收敛因子的非线性化引入灰狼算法从而形成一种改进的灰狼算法;然后,采用改进后的灰狼算法对KELM的惩罚系数和核参数进行优化;最后,将优化后的KELM应用于天波雷达定位,使建立的KELM定位模型具有更高的预测精度和更强的泛化能力。实验结果显示,所提模型的预测结果与目标实测值基本一致,预测精度高于标准灰狼优化算法改进的KELM模型和解析法定位模型,为天波雷达定位提供了一种新的目标定位方法。
Aimed at the disadvantages of the lower azimuth resolution of the sky-wave radar and larger position error of traditional analytic algorithm,a new locating model using chaotic mutation grey wolf optimization algorithm to optimize the kernel extreme learning machine(KELM)is put forward.First,the piecewise linear chaotic map,adaptive Cauchy mutation strategy and non-linearity of the convergence factor are introduced into the grey wolf optimization algorithm to form an improved grey wolf algorithm.Then,the improved grey wolf optimization algorithm is used to optimize penalty coefficient and kernel parameter of the KELM.Finally,the optimized the KELM is applied to sky-wave radar location,making the established KELM model have the high steady-state prediction accuracy and generalization performance.The experimental results show that the predicted results of the proposed model are basically consistent with the measured values,and the prediction accuracy is higher than that of the KELM location model,which is optimized by the standard grey wolf algorithm.A new target location method is provided for sky-wave radar.
作者
宋萍
刘以安
Song Ping;Liu Yian(College of IOT Engineering,Jiangnan University,Wuxi,Jiangsu 214122,China)
出处
《激光与光电子学进展》
CSCD
北大核心
2019年第3期234-242,共9页
Laser & Optoelectronics Progress
基金
国家自然科学基金(61170120)
关键词
光计算
天波雷达
定位
核极限学习机
灰狼优化算法
参数优化
optics in computing
sky-wave radar
location
kernel extreme learning machine
grey wolf optimization algorithm
parameter optimization
