th.2021.04.021]
点击复制

基于Kriging-KPLS的直流充电桩现场检测自适应补偿方法研究

《工业仪表与自动化装置》[ISSN:1000-0682/CN:61-1121/TH]

Title:: Research on adaptive compensation method of DC charging pile field verification based on Kriging-KPLS

作者:: 李贺龙1; 王春妍1; 徐静2; 梁航2; 赵进全2; 1.中国电力科学研究院有限公司，北京 100192；
2.西安交通大学电气工程学院，陕西西安 710049

Author(s):: LI Helong1; WANG Chunyan1; XU Jing2; LIANG Hang2; ZHAO Jinquan2; 1.Electrica Power Research Institute, Beijing 100192, China；
2.School of Electrical Engineering, Xi’an Jiaotong University,Shaanxi Xi’an 710049, China.

Keywords:: DC charging pile; on-site verification; Kriging-KPLS method; adaptive compensation

摘要:: 针对直流充电桩电能计量精度现场检定存在受现场工况影响大，检定装置检定结果不准确的问题，提出一种基于Kriging-KPLS的直流充电桩现场检定自适应补偿方法。该方法利用Kriging插值法，充分利用含有多维影响因素的电压电流误差实验数据点间的相关性及结构特征，对有限实验条件下的实验数据进行扩充，为建立准确的电压电流误差模型提供足够的数据量；利用KPLS方法将电压电流的Kriging扩充数据映射到高维特征空间中，在高维特征空间进行偏最小二乘回归，建立电压电流的误差模型并对电压电流进行在线补偿，减小了现场工况对检定装置的影响，提高了检定装置的抗干扰性能。通过实测验证了该文方法的有效性。

Abstract:: In view of the problems that the remote measurement of the DC charging pile energy measurement accuracy is greatly affected by the site conditions and the measurement results of the verification device are inaccurate,this paper proposes a Kriging-KPLS-based DC charging pile field verification anti-interference method.This method uses the Kriging interpolation method to fully consider the correlation and structural characteristics of the voltage and current error experimental data points containing multi-dimensional influencing factors,and expands the experimental data under limited experimental conditions to provide sufficient for establishing an accurate voltage and current error model.The amount of data use the KPLS method to map the Kriging extended data of voltage and current to the high-dimensional feature space,perform partial least squares regression in the high- dimensional feature space,establish an error model of voltage and current, and correct the voltage and current online to eliminate.The influence of environmental interference on the verification device is improved,and the anti-interference performance of the verification device is improved.The effectiveness of the method in this paper is verified by actual measurement.

[1]刘洋.国家电网建成全球接入充电桩数量最多的智慧车联网平台[J].电力系统装备,2019(21):24.

[2] 曾博,李刚,韩帅,等.电动汽车充电桩直流电能表检定系统[J].中国计量,2014(7):80-82.

[3] 全国电磁计量技术委员会.电动汽车非车载充电机:JJG 1149—2018[S].

[4] 李恺,杨茂涛,王智,等.充电桩直流计量模块现场检测装置及其溯源[J].电子测量技术,2017,40(09):265-270.

[5] 叶健诚,桑林,李昱凌,等.电动汽车非车载充电机谐波分析与仿真[J].电测与仪表,2011,48(09):34-37.

[6] 张靖.基于克里金算法的点云数据插值研究[D].西安:长安大学,2014.

[7] 吴雪梅,刘志强,张天龙,等.双模型结合进一步降低预测均方根误差和均方根相对误差的方法[J].分析化学,2015,43(05):754-758.

[8] 吕明东.考虑温度影响的智能电表计量精度一致性分析与优化[D].哈尔滨:哈尔滨工业大学,2018.

[9] 蒋红卫,夏结来,张春霞,等.核偏最小二乘回归及其在医学中的应用[J].中国卫生统计,2007(03):239-242.

[10] JOHN Shawe-Taylor,NELLO Cristianini.模式分析的核方法 [M].北京:机械工业出版社, 2006.

[11] 欣怡.基于核偏最小二乘回归的短期电力负荷预测研究[D].兰州:兰州交通大学,2017.

[12] 丁二锐,曾平,刘瑞华,等.基于提升核偏最小二乘回归的色彩校正[J].仪器仪表学报,2008(01):80-84.

[13] 中国电力企业联合会.电动汽车传导充电系统：GBT 18487.1—2015[S].

[14] 宋鹏,潘峰,林国营,等.电动汽车充电纹波对直流电能计量影响分析[J].电测与仪表,2018,55(01):46-52.

更新日期/Last Update: 1900-01-01