[1]徐彬泰,孟祥鹿,田安琪,等.基于粒子群优化及高斯过程回归的铅酸电池荷电状态预测[J].南京理工大学学报(自然科学版),2018,42(02):162.[doi:10.14177/j.cnki.32-1397n.2018.42.02.005]
 Xu Bintai,Meng Xianglu,Tian Anqi,et al.Prediction for state of charge of lead-acid battery by particleswarm optimization with Gaussian process regression[J].Journal of Nanjing University of Science and Technology,2018,42(02):162.[doi:10.14177/j.cnki.32-1397n.2018.42.02.005]
点击复制

基于粒子群优化及高斯过程回归的铅酸电池荷电状态预测()
分享到:

《南京理工大学学报》(自然科学版)[ISSN:1005-9830/CN:32-1397/N]

卷:
42卷
期数:
2018年02期
页码:
162
栏目:
出版日期:
2018-04-30

文章信息/Info

Title:
Prediction for state of charge of lead-acid battery by particleswarm optimization with Gaussian process regression
文章编号:
1005-9830(2018)02-0162-07
作者:
徐彬泰孟祥鹿田安琪孙勇健曹立斌江颖洁
国网山东省电力公司 信息通信公司,山东 济南 250001
Author(s):
Xu BintaiMeng XiangluTian AnqiSun YongjianCao LibinJiang Yingjie
Information & Telecommunication Company,State Grid Shandong ElectricPower Company,Jinan 250001,China
关键词:
铅酸电池荷电状态 高斯过程回归 粒子群优化 超参数优化
Keywords:
state of charge of lead-acid batteries Gaussian process regression particle swarm optimization hyperparameter optimization
分类号:
TM912.1
DOI:
10.14177/j.cnki.32-1397n.2018.42.02.005
摘要:
为了提高铅酸电池荷电状态(SOC)的预测准确率,该文提出一种基于粒子群优化的高斯过程回归(PSO-GPR)算法。该算法的核心思想是通过粒子群优化(PSO)算法来解决高斯过程回归(GPR)模型中的超参数优化问题。PSO-GPR首先随机生成一个粒子群,群中的每个粒子包含对应的GPR超参数信息。随后执行如下迭代步骤:根据当前每个粒子的超参数信息训练对应的GPR模型并评估该模型的性能,结合适应度函数和每个模型的评估结果计算出每个粒子的适应度,并更新每个粒子中的超参数信息; 经过多次迭代后,找到粒子群中适应度最小的粒子; 最后从该粒子中提取相应的超参数信息,并训练最终的GPR预测模型。在铅酸电池数据集上的实验结果表明,所提方法优于对比模型。
Abstract:
To improve the prediction accuracy of state of charge(SOC)for lead-acid batteries,a particle swarm optimization with Gaussian process regression(PSO-GPR)is designed here. The basic idea of PSO-GPR is using the particle swarm optimization(PSO)to optimize the hyperparameters of the Gaussian process regression(GPR). Firstly,the PSO-GPR randomly initializes several particles,and every parlicle contains the hyperparameters of GPR. Then the following iterations are executed:for each particle,the corresponding GPR is trained and evaluated by its information of hyperparameters; the fitness function is combined with the evaluation result of GPR to calculate the fitness of each particle,and then the hyperparameter information of all particles is updated. After multiple iterations,the particle sharing the lowest fitness is chosen,and the corresponding hyperparameters are extracted to train the final GPR. The experiment results in lead-acid battery datasets demonstrat that the proposed PSO-GPR outperforms other comparison models and shares the broad prospects.

参考文献/References:

[1] Qin Taichun,Zeng Shengkui,Guo Jianbin. Robust prognostics for state of health estimation of lithium-ion batteries based on an improved PSO-SVR model[J]. Microelectronics Reliability,2015,55(9-10):1280-1284.
[2]Burgos Claudio,Sáez Doris,Orchard Marcos E,et al. Fuzzy modelling for the state-of-charge estimation of lead-acid batteries[J]. Journal of Power Sources,2015,274(1):355-366.
[3]Li Zhi,Zhang Peng,Wang Zhifu,et al. State of charge estimation for Li-ion battery based on extended Kalman filter[J]. Energy Procedia,2017,105:3515-3520.
[4]Wang Weizhong,Wang Deqiang,Wang Xiao,et al. Comparison of Kalman filter-based state of charge estimation strategies for Li-ion batteries[C]//Proceedings of the IEEE Transportation Electrification Conference and Expo. Dearborn,US:ITEC Press,2016.
[5]Ozcan G,Pajovic M,Sahinoglu Z,et al. Online state of charge estimation for lithium-ion batteries using Gaussian process regression[C]//Proceedings of the 42nd Annual Conference of Industrial Electronics Society. Beijing,China:IECON Press,2016.
[6]赖少发,刘华军. 机动目标跟踪支持向量回归学习新方法[J]. 南京理工大学学报,2017,41(2):264-268.
Lai Shaofa,Liu Huajun. Novel approach in maneuvering target tracking based on support vector regression[J]. Journal of Nanjing University of Science and Technology,2017,41(2):264-268.
[7]门志国,彭秀艳,王兴梅,等. 基于GA优化BP神经网络辨识的Volterra级数核估计算法[J]. 南京理工大学学报,2012,36(6):962-967.
Men Zhiguo,Peng Xiuyan,Wang Xingmei,et al. Volterra series kernels estimation algorithm based on GA optimized BP neural network identification[J]. Journal of Nanjing University of Science and Technology,2012,36(6):962-967.
[8]王健峰,张磊,陈国兴,等. 基于改进的网格搜索法的SVM参数优化[J]. 应用科技,2012(3):28-31.
Wang Jianfeng,Zhang Lei,Chen Guoxing,et al. A parameter optimization method for an SVM based on improved grid search algorithm[J]. Applied Science and Technology,2012(3):28-31.
[9]孙斌,姚海涛,刘婷. 基于高斯过程回归的短期风速预测[J]. 中国电机工程学报,2012,32(29):104-109.
Sun Bin,Yao Haitao,Liu Ting. Short-term wind speed forecasting based on Gaussian process regression model[J]. Proceedings of the CSEE,2012,32(29):104-109.
[10]Genton M G. Classes of kernels for machine learning:a statistics perspective[J]. Journal of Machine Learning Research,2001,2(2):299-312.
[11]Stacy E W. A generalization of the gamma distribution[J]. Annals of Mathematical Statistics,1962,33(3):1187-1192.
[12]Andrews L C. Special functions of mathematics for engineers(Second Edition)[M]. New York,US:McGraw-Hill,1998.
[13]Long Bing,Xian Weiming,Jiang Lin,et al. An improved autoregressive model by particle swarm optimization for prognostics of lithium-ion batteries[J]. Microelectronics Reliability,2013,53(6):821-831.
[14]Efron B,Gong G. A leisurely look at the bootstrap,the jackknife,and cross-validation[J]. American Statistician,1983,37(1):36-48.

相似文献/References:

[1]张孙力,杨慧中.一种基于改进扩张搜索聚类算法的软测量建模方法[J].南京理工大学学报(自然科学版),2017,41(05):574.[doi:10.14177/j.cnki.32-1397n.2017.41.05.006]
 Zhang Sunli,Yang Huizhong.Soft sensor modeling method based on improved expandingsearching clustering algorithm[J].Journal of Nanjing University of Science and Technology,2017,41(02):574.[doi:10.14177/j.cnki.32-1397n.2017.41.05.006]

备注/Memo

备注/Memo:
收稿日期:2017-12-29 修回日期:2018-02-12
作者简介:徐彬泰(1987-),男,硕士,工程师,主要研究方向:电力通信、无线通信,E-mail:xubintai123@163.com; 通讯作者:江颖洁(1989-),女,硕士,助理工程师,主要研究方向:通信线路运检,E-mail:asas940221@163.com。
引文格式:徐彬泰,孟祥鹿,田安琪,等. 基于粒子群优化及高斯过程回归的铅酸电池荷电状态预测[J]. 南京理工大学学报,2018,42(2):162-168.
投稿网址:http://zrxuebao.njust.edu.cn
更新日期/Last Update: 2018-04-30