作者：郁美霞1, 董刚1, 胥如迅2, 喻国丽1
作者单位：1. 甘肃省计量研究院, 甘肃 兰州 730050;
2. 兰州交通大学机电技术研究所, 甘肃 兰州 730070
关键词：精密时基源;BP神经网络;预测模型;归一化
摘要：
针对精密时基源溯源校准周期固定、效率低，提出基于BP神经网络构建精密时基源校准预测模型，可为溯源校准提供新方法。首先分析精密时基源校准参数产生机理及预测目标的影响因素，分别构建时间驱动和数据驱动校准预测模型；其次基于BP神经网络学习规则，归一化处理训练及预测数据，规避数据特征、分布差异对预测模型的不良影响；最后根据预测目标选择不同校准预测模型。通过仿真分析表明，时间驱动预测精度优于数据驱动预测精度，且两种预测模型预测误差均在10–10量级上，满足预测精度要求。
Calibration prediction model of precision time base source based on neural network
YU Meixia, DONG Gang, XU Ruxun, YU Guoli
1. Gansu Institute of Metrology, Lanzhou 730050, China;
2. Mechatronics T & R Institute, Lanzhou Jiaotong University, Lanzhou 730070, China
Abstract: For the low efficiency and fixed period of precision time base calibration, a prediction model of precision time base calibration based on BP neural network is proposed to provide a new method for traceability calibration. Firstly, the prediction target influencing factors and calibration parameters generation mechanism of precision time base source are analyzed, and the time driven and data driven calibration prediction models are established. Secondly, based on BP neural network learning rules, the training and prediction data are normalized to avoid the adverse effects of distribution differences and data characteristics on the prediction model. Finally, different calibration prediction models are selected according to the prediction target. The simulation analysis shows that, the accuracy of time driven model is better than that of data driven model, and the prediction errors of the two prediction models are on the order of 10–10, which meets the prediction accuracy requirements.
Keywords: precision time base source;BP neural network;prediction model;normalization
2023, 49(9):194-200 收稿日期: 2022-7-6;收到修改稿日期: 2022-10-24
基金项目: 甘肃省青年科技基金（23JRRA1270）；甘肃省市场监督管理局科技计划项目(SSCJG-JL-202101)
作者简介: 郁美霞（1989-），女，山东莒南县人，工程师，硕士，主要从事时间频率计量及状态预测方面的研究。
参考文献
[1] 杨枫. 支持交会对接任务精密时间基准方法研究与实现[J]. 计算机测量与控制, 2020, 28(11): 145-149.
[2] 时钟测试仪校准规范: JJF1662—2017[S]. 北京: 中国质检出版社, 2017.
[3] 肖化, 胡广莉, 何惠玲, 等. 基于分组BP神经网络的两相流电容层析技术[J]. 计量学报, 1998(3): 50-54.
[4] 刘通, 秦正山, 何芳, 等. 基于神经网络的磁性活性炭吸附双酚A研究[J]. 磁性材料及器件, 2019, 50(1): 9-11.
[5] 陈春俊, 杨露, 何智颖, 等. ARIMA-BP神经网络高速列车隧道压力波预测模型研究[J]. 中国测试, 2021, 47(10): 80-86.
[6] 赵锐, 赵丽萍, 陈静芳, 等. 基于BP神经网络的垃圾渗滤液输运管道结垢趋势预测[J]. 中国测试, 2022, 48(7): 1-7.
[7] 王伟, 李永恒, 李思琪, 等. 基于改进BP神经网络的钢筋混凝土井壁极限承载力随机分析[J/OL]. 中国测试: 1-6[2022-06-16]. http://kns.cnki.net/kcms/detail/51.1714.TB.20220628.1106.006.html.
[8] WEN F, JING F S, ZHAO W H, et al. Research on optimal receiver radius of wireless power transfer system based on BP neural network[J]. Energy Reports, 2020, 6(9): 1450-1455.
[9] FENG J Y, YUAN B Y, LI X, et al. Evaluation on risks of sustainable supply chain based on optimized BP neural networks in fresh grape industry[J]. Computers and Electronics in Agriculture, 2021, 183: 105988.
[10] YU L H, XIE L Y, LIU C M, et al. Optimization of BP neural network model by chaotic krill herd algorithm[J]. Alexandria Engineering Journal, 2022, 61(12): 9769-9777.
[11] 南国芳, 王化祥, 王超. 基于BP神经网络和遗传算法的电阻抗图像重建算法[J]. 计量学报, 2003(4): 337-340.
[12] 王彦明, 贾克斌, 刘鹏宇, 等. 基于增强型BP网络的气象传感器标校方法[J]. 中国测试, 2020, 46(12): 105-111.
[13] 王敏, 邹滨, 郭宇, 等. 基于BP人工神经网络的城市PM2.5浓度空间预测[J]. 环境污染与防治, 2013, 35(9): 63-66+70.
[14] 谢胜龙, 张文欣, 张为民, 等. 基于多分支BP神经网络的气动肌肉迟滞建模方法[J]. 计量学报, 2021, 42(6): 745-752.
[15] ZHANG R, WANG Y, WANG K, et al. An evaluating model for smart growth plan based on BP neural network and set pair analysis[J]. Journal of Cleaner Production, 2019, 226: 928-939.
[16] XU Y, HUANG Y M, MA G W. A beetle antennae search improved BP neural network model for predicting multi-factor-based gas explosion pressures[J]. Journal of Loss Prevention in the Process Industries, 2020, 65: 104117.
[17] SONG S H, XIONG X Y, WU X, et al. Modeling the SOFC by BP neural network algorithm[J]. International Journal of Hydrogen Energy, 2021, 46(38): 20065-20077.
[18] ZHOU Z, LI Y F, LI F M, et al. An intelligence energy consumption model based on BP neural network in mobile edge computing[J]. Journal of Parallel and Distributed Computing, 2022, 167: 211-220.
