Please wait a minute...
浙江大学学报(工学版)
浙江大学学报(工学版)  2025, Vol. 59 Issue (7): 1547-1556    DOI: 10.3785/j.issn.1008-973X.2025.07.023
航空航天技术     
高精度测量装置大气环境参数校准深度学习混合代理模型
袁新博1(),徐兆斌1,2,*(),李自茹1,潘健1,金小军1,2,3,金仲和1,2,3
1. 浙江大学 微小卫星研究中心，浙江 杭州 310027
2. 浙江省微纳卫星研究重点实验室，浙江 杭州 310027
3. 浣江实验室，浙江 诸暨 311899
Deep learning hybrid agent model for atmospheric environment parameter calibration in high-precision measurement devices
Xinbo YUAN1(),Zhaobin XU1,2,*(),Ziru LI1,Jian PAN1,Xiaojun JIN1,2,3,Zhonghe JIN1,2,3
1. Micro-Satellite Research Center, Zhejiang University, Hangzhou 310027, China
2. Key Laboratory of Micro-Nano Satellite Research of Zhejiang Province, Hangzhou 310027, China
3. Huanjiang Laboratory, Zhuji 311899, China
 全文: PDF(1398 KB)   HTML
摘要：

为了提高测量装置在野外校准时的大气折射率补偿精度，提出基于深度学习理论的高精度物理场混合代理拟合模型. 正逆向时间序列预测扩充传感器阵列，获取高精度的环境参数优化大气折射补偿系统，增加同一时刻传感器阵列观测点位的分布密集程度. 以气象参数中的温度场为例，仿真结果表明，所提模型相较于传统多点法补偿模型具有更高的精度，所提模型使环境参数的精度提高了71.8%，标准差降低了73.1%. 蒙特卡洛仿真分析的结果证明，所提模型相较于传统多点法具有更强的稳定性.
关键词： 测量装置;  环境参数;  传感器阵列;  时间序列分析;  径向基函数插值    
Abstract:

A high-precision physical field hybrid agent fitting model based on deep learning theory was proposed, in order to improve the compensation accuracy of a measurement device for the atmospheric refraction index in-field calibration. The sensor array was expanded by forward and backward time series prediction, the atmospheric refraction compensation system was optimized by obtaining high-accuracy environmental parameters, and the density distribution of sensor array observation points was increased at the same moment. Taking the temperature field among meteorological parameters as an example, simulation results show that the proposed model achieves higher accuracy than traditional multi-point compensation models. The proposed model improved the accuracy of environmental parameters by 71.8% and reduced the standard deviation by 73.1%. Monte Carlo simulation analysis demonstrates that the proposed model exhibits stronger stability compared to traditional multi-point methods.
Key words: measurement device    environmental parameter    sensor array    time series analysis    radial basis function interpolation
收稿日期: 2024-06-26 出版日期: 2025-07-25
CLC:  TN 98  
基金资助: 国家自然科学基金资助项目（U21A20443，62073289）.
通讯作者: 徐兆斌     E-mail: xinbo@zju.edu.cn;zjuxzb@zju.edu.cn
作者简介: 袁新博（2000—），男，硕士生，从事高精度测量、卫星路由研究. orcid.org/0009-0000-4802-5408. E-mail：xinbo@zju.edu.cn
服务　　
把本文推荐给朋友
加入引用管理器
E-mail Alert
作者相关文章　　
袁新博
徐兆斌
李自茹
潘健
金小军
金仲和

引用本文:

袁新博,徐兆斌,李自茹,潘健,金小军,金仲和. 高精度测量装置大气环境参数校准深度学习混合代理模型[J]. 浙江大学学报(工学版), 2025, 59(7): 1547-1556.

Xinbo YUAN,Zhaobin XU,Ziru LI,Jian PAN,Xiaojun JIN,Zhonghe JIN. Deep learning hybrid agent model for atmospheric environment parameter calibration in high-precision measurement devices. Journal of ZheJiang University (Engineering Science), 2025, 59(7): 1547-1556.

链接本文:

https://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2025.07.023    或    https://www.zjujournals.com/eng/CN/Y2025/V59/I7/1547

图 1  典型一维卷积神经网络架构图
图 2  激活函数性能对比图
图 3  基于深度学习的扩充传感器阵列模型
图 4  逆向时间序列分析示意图
图 5  基于正逆向时间序列扩充传感器阵列的模型
图 6  高次插值导致的龙格现象
图 7  融合径向基函数插值与多项式拟合的算法框架
图 8  测距环境仿真模型
图 9  扩充传感器阵列的仿真对比图
图 10  扩充传感器阵列前后的平均温度偏差对比
dlog (RMSE)dlog (RMSE)
5?3.0411?3.18
7?3.4113?1.39
9?3.47——
表 1  不同多项式次数拟合的对数均方根误差
图 11  不同插值方法的物理场
图 12  多项式拟合时的损失函数（以均方根误差为损失函数）
图 13  经自适应正则化的拟合损失函数
图 14  多项式拟合时正则化的效果图
图 15  蒙特卡洛法仿真的正态分布分析
方法$\mu $$\sigma $
多点法$ 1.71 \times {10^{ - 2}} $$1.31 \times {10^{ - 2}}$
时间序列+多点法$5.44 \times {10^{ - 3}}$$ 4.01 \times {10^{ - 3}} $
时间序列+插值拟合$4.82 \times {10^{ - 3}}$$3.53 \times {10^{ - 3}}$
表 2  正态分布的均值及标准差
图 16  传感器补偿系统示意图
图 17  测线内的温湿度变化监测结果
图 18  室内实验平台
$d_{\mathrm{r}}$/m$\Delta {d_{{\text{pre}}}}$/mm$\Delta {d_{{\text{cor}}}}$/μm
102.986.54
205.996.83
308.938.53
4011.887.02
5014.8711.62
表 3  修正前后的测距偏差
1 李萍, 李建双, 赫明钊, 等 基于空气环境参数自动测量系统的全站仪测距精度校准技术[J]. 计量学报, 2019, 40 (6A): 12- 16
LI Ping, LI Jianshuang, HE Mingzhao, et al Ranging accuracy calibration technology of total station based on automatic measurement system of air environment parameters[J]. Acta Metrologica Sinica, 2019, 40 (6A): 12- 16
doi: 10.3969/j.issn.1000-1158.2019.z1.03
2 叶孝佑, 甘晓川, 常海涛, 等 提高环境参量法测量空气折射率精度的研究[J]. 计量学报, 2008, 29 (4A): 69- 73
YE Xiaoyou, GAN Xiaochuan, CHANG Haitao, et al The improvement of measuring precision for refractive index of air using the method of Edlén equation calculating with environmental parameters of air[J]. Acta Metrologica Sinica, 2008, 29 (4A): 69- 73
3 AZOUIGUI S, BADR T, WALLERAND J P, et al Transportable distance measurement system based on superheterodyne interferometry using two phase-locked frequency-doubled Nd: YAG lasers[J]. Review of Scientific Instruments, 2010, 81 (5): 053112
doi: 10.1063/1.3428729
4 张海波, 张健, 韩正阳 野外基线温度测量方法探讨[J]. 工业计量, 2017, 27 (1): 36- 37
ZHANG Haibo, ZHANG Jian, HAN Zhengyang Exploration of methods for measuring baseline temperature in the field[J]. Industrial Metrology, 2017, 27 (1): 36- 37
5 POLLINGER F, MEYER T, BEYER J, et al The upgraded PTB 600 m baseline: a high-accuracy reference for the calibration and the development of long distance measurement devices[J]. Measurement Science and Technology, 2012, 23 (9): 094018
doi: 10.1088/0957-0233/23/9/094018
6 陈杨, 李建双, 缪东晶, 等 基于传感器阵列的野外基线环境参数自动测量系统研制[J]. 计量学报, 2018, 39 (4): 455- 460
CHEN Yang, LI Jianshuang, MIAO Dongjing, et al Development of outdoor baseline environment parameters measurement system based on sensor array[J]. Acta Metrologica Sinica, 2018, 39 (4): 455- 460
doi: 10.3969/j.issn.1000-1158.2018.04.02
7 刘学德, 缪东晶, 张京燕, 等 1.2 km标准基线环境参数自动测量系统研制[J]. 计量学报, 2020, 41 (8): 897- 902
LIU Xuede, MIAO Dongjing, ZHANG Jingyan, et al Development of environmental parameters automatic measurement system for 1.2 km standard baseline[J]. Acta Metrologica Sinica, 2020, 41 (8): 897- 902
doi: 10.3969/j.issn.1000-1158.2020.08.01
8 谷友艺, 蒋理兴, 王力, 等 野外基线测温方法的比较研究[J]. 测绘与空间地理信息, 2019, 42 (12): 27- 30
GU Youyi, JIANG Lixing, WANG Li, et al Comparative study on thermometry methods of outdoor baseline[J]. Geomatics and Spatial Information Technology, 2019, 42 (12): 27- 30
9 LIU G, ZHONG K, LI H, et al A state of art review on time series forecasting with machine learning for environmental parameters in agricultural greenhouses[J]. Information Processing in Agriculture, 2024, 11 (2): 143- 162
doi: 10.1016/j.inpa.2022.10.005
10 TORRES J F, HADJOUT D, SEBAA A, et al Deep learning for time series forecasting: a survey[J]. Big Data, 2021, 9 (1): 3- 21
doi: 10.1089/big.2020.0159
11 FIME A A, ASHIKUZZAMAN M, AZIZ A Audio signal based danger detection using signal processing and deep learning[J]. Expert Systems with Applications, 2024, 237: 121646
doi: 10.1016/j.eswa.2023.121646
12 YANG R, SINGH S K, TAVAKKOLI M, et al CNN-LSTM deep learning architecture for computer vision-based modal frequency detection[J]. Mechanical Systems and signal processing, 2020, 144: 106885
doi: 10.1016/j.ymssp.2020.106885
13 ZHOU Y, LI J, CHI J, et al Set-CNN: a text convolutional neural network based on semantic extension for short text classification[J]. Knowledge-Based Systems, 2022, 257: 109948
doi: 10.1016/j.knosys.2022.109948
14 BHATT D, PATEL C, TALSANIA H, et al CNN variants for computer vision: history, architecture, application, challenges and future scope[J]. Electronics, 2021, 10 (20): 2470
doi: 10.3390/electronics10202470
15 金列俊, 詹建明, 陈俊华, 等 基于一维卷积神经网络的钻杆故障诊断[J]. 浙江大学学报: 工学版, 2020, 54 (3): 467- 474
JIN Liejun, ZHAN Jianming, CHEN Junhua, et al Drill pipe fault diagnosis method based on one-dimensional convolutional neural network[J]. Journal of Zhejiang University: Engineering Science, 2020, 54 (3): 467- 474
16 金海龙, 马吴旭, 孟宗, 等 基于改进1DCNN-GRU的滚动轴承故障诊断[J]. 计量学报, 2023, 44 (9): 1423- 1428
JIN Hailong, MA Wuxu, MENG Zong, et al An improved 1DCNN-GRU for rolling bearing fault diagnosis[J]. Acta Metrologica Sinica, 2023, 44 (9): 1423- 1428
doi: 10.3969/j.issn.1000-1158.2023.09.16
17 SHE J, TAN Y Research on Runge phenomenon[J]. Advances in Applied Mathematics, 2019, 8 (8): 1500- 1510
doi: 10.12677/AAM.2019.88175
18 MAZARGUIL A, OUDRE L, VAYATIS N Non-smooth interpolation of graph signals[J]. Signal Processing, 2022, 196: 108480
doi: 10.1016/j.sigpro.2022.108480
19 YU Z, SONG D, SONG Y, et al Research on a visual thermal landscape model of underground space based on the spatial interpolation method: a case study in Shanghai[J]. Energy Reports, 2022, 8: 406- 418
20 CHEN J, CAI Z Cardinal MK-spline signal processing: spatial interpolation and frequency domain filtering[J]. Information Sciences, 2019, 495: 116- 135
doi: 10.1016/j.ins.2019.04.056
21 郭晓旭, 徐兆斌, 成恒飞, 等 用于星间通信的高灵敏度快捕算法[J]. 哈尔滨工业大学学报, 2023, 55 (8): 43- 50
GUO Xiaoxu, XU Zhaobin, CHENG Hengfei, et al High sensitivity fast acquisition algorithm for inter-satellite communication[J]. Journal of Harbin Institute of Technology, 2023, 55 (8): 43- 50
doi: 10.11918/202208112
[1] 胡比澜,王洋洋,张永强. 基于多模型Stacking融合的基坑测斜时序预测[J]. 浙江大学学报(工学版), 2025, 59(4): 706-716.
[2] 陈功,郑春华,翁献明,HAMEEDBaustani,胡鸿昊,马晓宇,柳景青. 基于多水文数据关联分析的雨水口异常诊断[J]. 浙江大学学报(工学版), 2021, 55(1): 55-61.
[3] 柯显信, 张文朕, 杨阳, 温雷. 仿人机器人多传感器定位系统[J]. 浙江大学学报(工学版), 2018, 52(7): 1247-1252.
[4] 魏媛,冯天恒,黄平捷,侯迪波,张光新. 管网水质多指标动态关联异常检测方法[J]. 浙江大学学报(工学版), 2016, 50(7): 1402-1409.
[5] 付兴伟, 吴功平, 周鹏, 于娜. 基于卡尔曼滤波的巡视机器人能耗估计[J]. 浙江大学学报(工学版), 2015, 49(4): 670-675.
[6] 王智磊,孙红月,刘永莉,尚岳全. 降雨与边坡地下水位关系的时间序列分析[J]. J4, 2011, 45(7): 1301-1307.