In this paper, a new learning algorithm called variable activation function extreme learning machine based on residual prediction compensation (RV-ELM) is proposed to solve the problem that ELM algorithm uses fixed activation function and has not residual compensation. In the learning process, it uses particle swarm optimization algorithm to optimize the steep degree, position and mapping scope simultaneously, and the ARMA model is used to model the residual errors between actual value and prediction value of variable activation function extreme learning machine (V-ELM), and the prediction of residual errors is used to rectify the prediction value of V-ELM. Simulation experiment is performed to show the effectiveness and feasibility of this method for benchmark datasets. And it is utilized to develop a soft sensor model for the surface-to-air missile survivability, and the result was satisfied.
WANG Gaitang
,
WANG Honghui
,
ZHAO Jinlei
,
BAI Yanfang
. Variable Activation Function Extreme Learning Machine Based on Residual Prediction Compensation[J]. Journal of Projectiles, Rockets, Missiles and Guidance, 2017
, 37(6)
: 129
-132
.
DOI: 10.15892/j.cnki.djzdxb.2017.06.030
[an error occurred while processing this directive]
[1] HUANG Guangbin, ZHU Qinyu, SIEW Chee-Kheong. Extreme learning machine: Theory and applications[J].Neurocomputing, 2006, 70(1/2/3): 489-501.
[2] HUYNH H T, WON Y. Extreme learning machine with fuzzy activation function[C]//IEEE. Proceedings of 2009 Fifth International Joint Conference on INC, IMS and IDC.[S.l.]: IEEE, 2009: 303-307.
[3] LIANG N Y, HUANG G B, SARATCHANDRAN N, et al. A fast and accurate online sequential learning algorithm for feedforward networks[J].IEEE Tansactions on Neural Networks,2006, 17(6) : 1411-1423.
[4] LAN Yuan, SOH Yeng Chai, HUANG Guang-Bin. Ensemble of online sequential extreme learning machine[J].Neurocomputing, 2009, 72(13/14/15): 3391-3395.
[5] HUANG Guang-Bin, CHEN Lei. Convex incremental extreme learning machine[J].Neurocomputing, 2007, 70(16/17/18): 3056-3062.
[6] 张文博, 姬红兵, 王磊,等. 多隐层输出矩阵极限学习机[J].系统工程与电子技术, 2014, 36(8): 1656-1659.
[7] 徐圆, 卢玉帅, 才轶. 基于多元时序驱动的ELM全流程故障预测及其应用[J].化工学报, 2015, 66(1): 351-356.
[8] 颜七笙, 王士同, 张延, 等. 基于经验模式分解和极限学习机的铀资源价格预测方法[J].控制与决策, 2014, 29(7): 1187-1192.
[9] 宋绍剑, 向伟康, 林小峰. 增量型极限学习机改进算法[J].信息与控制, 2016, 45(6): 735-741.
[10] 张文博, 姬红兵. 融合极限学习机[J].电子与信息学报, 2013, 35(11): 2728-2732.
[11] WANG Gaitang, LI Ping, SU Chengli. Soft sensor modeling based on multiple extreme learning machine[J].Advanced Materials Research, 2012, 433-440: 3003-
3010.
[12] 王海燕, 卢山. 非线性时间序列分析及其应用[M].北京: 科学出版社, 2006: 45-68.
[13] 胡耀垓, 李伟, 胡继明. 一种改进激活函数的人工神经网络及其应用[J].武汉大学学报(信息科学版), 2004, 29(10): 916-919.
[14] KENNEDY J, EBERHERT R. Particle swarm optimization[J].IEEE. Proceedings of ICNN’95-International Conference on Neural Networks.[S.l.]:IEEE, 1995: 192-194.