基于改进RBF神经网络训练算法的蓄电池SOC估算SOC estimation of battery based on improved RBF neural network training algorithm
陈玮匀,杨文伟,陈俊江,胡永乐,覃团发
摘要(Abstract)：
为了对基站铅酸蓄电池的剩余容量进行估算,在RBF神经网络训练算法中采用梯度下降法结合L1,L2正则化实现隐层节点数的选择。通过模糊控制思想提高泛化能力,引入模拟退火算法减少了需要重复训练的次数。Matlab仿真结果表明,通过改进的RBF神经网络训练算法对蓄电池的电池剩余容量(SOC)估算平均误差达到2%,改进了估算精度的同时也提高了泛化能力并且减少了重复训练的次数。
关键词(KeyWords)： SOC估算;RBF神经网络;正则化;模糊控制;模拟退火算法;Matlab
基金项目(Foundation): 国家自然科学基金项目(61761007);; 广西自然科学基金项目(2016GXNSFAA380222)~~
作者(Author): 陈玮匀,杨文伟,陈俊江,胡永乐,覃团发
DOI: 10.16652/j.issn.1004-373x.2018.20.035
参考文献(References)：
- [1]潘成举.基于物联网的基站蓄电池运维及应急发电调度系统开发[D].南宁:广西大学,2015.PAN Chengju. Development of operation maintenance of base station′s battery and emergency generation dispatching system based on Internet of Things[D]. Nanning:Guangxi University,2015.
- [2]余滨杉,王社良,杨涛,等.基于遗传算法优化的SMABP神经网络本构模型[J].金属学报,2017,53(2):248-256.YU Binshan,WANG Sheliang,YANG Tao,et al. BP neural network constitutive model based on optimization with genetic algorithm for SMA[J]. Acta Metallurgica Sinica, 2017, 53(2):248-256.
- [3]曾谁飞,张笑燕,杜晓峰,等.基于神经网络的文本表示模型新方法[J].通信学报,2017,38(4):86-98.ZENG Shuifei,ZHANG Xiaoyan,DU Xiaofeng,et al. New method of text representation model based on neural network[J]. Journal on communications,2017,38(4):86-98.
- [4] WANG Qianqian,WANG Jiao,ZHAO Pengju,et al. Correlation between the model accuracy and model-based SOC estimation[J]. Electrochimica acta,2017,228:146-159.
- [5] XU C,LU C,LIANG X,et al. Multi-loss regularized deep neural network[J]. IEEE transactions on circuits and systems for video technology,2016,26(12):2273-2283.
- [6] TIAN Yuan,YU Yuanlong. A new pruning algorithm for extreme learning machine[C]//Proceedings of IEEE International Conference on Information&Automation. Macau:IEEE,2017:704-709.
- [7] SEVAKULA R K,VERMA N K. Assessing generalization ability of majority vote point classifiers[J]. IEEE transactions on neural networks&learning systems,2017,28(12):2985-2997.
- [8] LI X,ZHU Y,XIA P. Enhanced analog beamforming for single carrier millimeter wave MIMO systems[J]. IEEE transactions on wireless communications,2017,16(7):4261-4274.
- [9]孟凡超,初佃辉,李克秋,等.基于混合遗传模拟退火算法的SaaS构件优化放置[J].软件学报,2016,27(4):916-932.MENG Fanchao,CHU Dianhui,LI Keqiu,et al. Solving SaaS components optimization placement problem with hybrid genetic and simulated annealing algorithm[J]. Journal of software,2016,27(4):916-932.
- [10]李珂,顾欣,刘旭东,等.基于梯度下降法的永磁同步电机单电流弱磁优化控制[J].电工技术学报,2016,31(15):8-15.LI Ke,GU Xin,LIU Xudong,et al. Optimized flux weakening control of IPMSM based on gradient descent method with single-current regulator[J]. Transactions of China Electrotechnical Society,2016,31(15):8-15.