基于神经网络的扭矩传感器稳定性分析预测

孟晨; 蒋继乐; 郭斌; 吴鲲; 吴实

doi:10.12338/j.issn.2096-9015.2021.0633

基于神经网络的扭矩传感器稳定性分析预测

doi: 10.12338/j.issn.2096-9015.2021.0633

中国计量科学研究院，北京 100029

基金项目: 国家重点研发计划资助项目（2017YFF0204905）；国家市场监督管理总局质量技术基础能力建设项目(ANL2009)；中国计量科学研究院基本科研业务费重点领域项目（AKYZD2005-1）。

详细信息

作者简介:
孟晨（1993-），中国计量科学研究院助理工程师，研究方向：力值计量、算法分析，邮箱：mengchen@nim.ac.cn

通讯作者:
蒋继乐（1984-），中国计量科学研究院副研究员，研究方向：力值扭矩计量，邮箱：jiangjl@nim.ac.cn

计量
- 文章访问数: 450
- HTML全文浏览量: 206
- PDF下载量: 48
- 被引次数: 0
出版历程
- 录用日期: 2022-04-07
- 网络出版日期: 2022-06-02
- 刊出日期: 2022-07-11

Research on Prediction of Stability of Torque Sensor Based on Neural Network

National Institute of Metrology, Beijing 100029, China

摘要

摘要: 在扭矩检定中，扭矩值随时间的变化波动并非是已知的常见函数，探究其变化波动趋势对于扭矩检定及量值传递有一定价值和意义。为了探究扭矩传感器示值波动特性，针对扭矩测量中长期稳定性对扭矩值变化的影响进行研究，采用神经网络对给定条件下的检测数据进行训练，对其他时间节点的扭矩值进行预测。以严格控制实验条件下得到的近12年的扭矩值数据进行训练和分析，采用交叉验证的方式证明了模型的准确性和可靠性，并对下一检定周期的扭矩值进行预测，最终计算得到下一周期扭矩值满足0.03级的长期稳定性的概率大于71.8%。
- 长期稳定性 /
- 神经网络 /
- 交叉验证 /
- 扭矩传感器
Abstract: In torque verification, the fluctuation of torque value over time is not a known common function, and it is vital to explore the trend of its change and fluctuation for torque verification and dissemination of value of quantity. To explore the fluctuation characteristic of the torque sensor, the influence of medium- and long-term stability on the change of torque value in torque measurement is investigated, and a neural network is used to train the detection data under given conditions to predict the torque value at other time points. The torque value data obtained under strictly controlled experimental conditions for the last 12 years were used for training and analysis, the accuracy and reliability of the model were demonstrated by cross-validation, the torque value of the next calibration cycle was predicted, and the probability that the torque value of the next cycle meets the long-term stability of 0.03 level was finally calculated to be greater than 71.8%.
- long-term stability /
- neural network /
- cross-validation /
- torque sensor

HTML全文

图 1 神经网络结构图

Figure 1. Neural network framework

下载: 全尺寸图片幻灯片

图 2 交叉验证对比结果

Figure 2. Cross validation comparison results

下载: 全尺寸图片幻灯片

图 3 交叉验证对比结果（相对值）

Figure 3. Cross-validation comparison results (relative value)

下载: 全尺寸图片幻灯片

图 4 扭矩输出预测结果

Figure 4. Prediction on readout of the torque sensor

下载: 全尺寸图片幻灯片

图 5 预测结果与真实值

Figure 5. Comparison of prediction value and

下载: 全尺寸图片幻灯片

图 6 恒温恒湿条件下的预测值

Figure 6. Predicted value under constant temperature and humidity

下载: 全尺寸图片幻灯片

表 1 扭矩实测结果

Table 1. Torque measurement results

载荷值	使用年月	温度（℃）/ 湿度（%RH）	包含区间最小值	包含区间最大值
100-CW	0	22/20	0.135011	0.135029
500-CW	0	22/20	0.675015	0.675043
1000-CW	0	22/20	1.350115	1.350166
100-CCW	0	22/20	−0.134977	−0.134999
500-CCW	0	22/20	−0.674945	−0.674976
1000-CCW	0	22/20	−1.350029	−1.350061

下载: 导出CSV

表 2 交叉验证对比结果

Table 2. Cross-validation comparison results

载荷值（N·m）	预测下限（mV/V）	真实值（mV/V）	预测上限（mV/V）
100-CW	0.134950	0.134951	0.135027
500-CW	0.674760	0.674879	0.675131
1000-CW	1.349613	1.349896	1.350322
100-CCW	−0.134955	−0.134996	−0.135033
500-CCW	−0.674799	−0.674954	−0.675202
1000-CCW	−1.349746	−1.350003	−1.350585

下载: 导出CSV

表 3 预测结果置信水平

Table 3. Confidence level of prediction

载荷值（N·m）	预测下限（mV/V）	上周期示值（mV/V）	预测上限（mV/V）	概率（%）
100-CW	0.134950	0.134951	0.135026	71.8
500-CW	0.674760	0.674879	0.675130	95.4
1000-CW	1.349613	1.349896	1.350322	95.4
100-CCW	−0.134954	−0.134996	−0.135033	92.7
500-CCW	−0.674799	−0.674954	−0.675202	95.4
1000-CCW	−1.349746	−1.350003	−1.350584	92.7

下载: 导出CSV

表 4 预测结果与真实测量

Table 4. Comparison of prediction value and measurement results

载荷值（N·m）	预测下限（mV/V）	真实值（mV/V）	预测上限（mV/V）
100-CW	0.134950	0.134983	0.135026
500-CW	0.674760	0.674884	0.675130
1000-CW	1.349613	1.349897	1.350322
100-CCW	−0.134954	−0.134994	−0.135033
500-CCW	−0.674799	−0.674950	−0.675202
1000-CCW	−1.349746	−1.350003	−1.350584

下载: 导出CSV

参考文献(19)

[1]	Anil K Jain, Jianchang Mao, K Moidin Mohiuddin. Artificial Neural Networks: A Tutorial[J]. Computer, 1996, 29(3): 31-44. doi: 10.1109/2.485891
[2]	Lippmann R P. An introduction to computing with neural nets[J]. IEEE Assp Magazine, 1988, 4(2): 4-22.
[3]	DE Rumelhart, Hinton G E, Williams R J. Learning Representations by Back Propagating Errors[J]. Nature, 1986, 323(6088): 533-536. doi: 10.1038/323533a0
[4]	全国力值硬度计量技术委员会. 标准扭矩仪: JJG 557-2011 [S]. 北京: 中国计量出版社, 2011.
[5]	ISO/TC. Metallic materials-Calibration of force-proving instruments used for the verification of uniaxial testing machines: ISO 376: 2011[S]. Geneva: International Standards Organization, 2011.
[6]	全国法制计量管理计量技术委员会. 测量不确定度评定与表示: JJF1059.1-2012[S]. 北京: 中国质检出版社, 2012.
[7]	Robert Hecht-Nielsen. Theory of the backpropagation neural network[C]. In Proceeding of the International Joint Conference on Neural Networks(IJCNN). 1989: 593-605.
[8]	Hornik K. Approximation capabilities of multilayer feedforward networks[J]. Neural Networks, 1991, 4(2): 251-257. doi: 10.1016/0893-6080(91)90009-T
[9]	Hornik K, Stinchcombe M, White H. Multilayer feedforward networks are universal Approximators[J]. Neural Networks, 1989, 2: 359-366. doi: 10.1016/0893-6080(89)90020-8
[10]	Cybenko G. Approximation by superpositions of a sigmoid function[J]. Mathematics of Control, Signals, and Systems, 1989, 2: 303-314. doi: 10.1007/BF02551274
[11]	Baratto-Filho F, Freitas J, Gabardo M C L, et al. Analysis of reciprocating system real-time torque variation during root canal preparation[J]. BMC Research Notes, 2019, 12(1): 47. doi: 10.1186/s13104-019-4080-z
[12]	Kwak S W, Ha J H, Cheung S P, et al. Comparison of In Vitro Torque Generation during Instrumentation with Adaptive Versus Continuous Movement[J]. Journal of Endodontics, 2019(45): 803-807.
[13]	Kwak S W, Ha J H, Cheung S P, et al. Effect of the Glide Path Establishment on the Torque Generation to the Files during Instrumentation An InVitro Measurement[J]. J Endod, 2018, 44(3): 496-500. doi: 10.1016/j.joen.2017.09.016
[14]	Bürklein S, Stüber JP, Schäfer E. Real-time dynamic torque values and axial forces during preparation of straight root canals using three different endodontic motors and hand preparation[J]. Int Endod J, 2019(52): 94-104.
[15]	Arvind Yadav, Boggavarapu Bhanu Venkata Satya Vara Prasad, Ramesh Kumar Mojjada, et al. Application of Artificial Neural Network and Genetic Algorithm Based Artificial Neural Network Models for River Flow Prediction[J]. Revue d'Intelligence Artificielle, 2020, 34(6): 745-751. doi: 10.18280/ria.340608
[16]	蒋继乐, 吴鲲, 郭斌, 等. 应用于惯性导航系统中的扭矩计量技术研究[J]. 计量科学与技术, 2021(2): 59-62,75.
[17]	张智敏, 张跃, 孟峰, 等. 基于空气轴承支撑技术的高准确度大扭矩标准装置研究[J]. 计量学报, 2021, 42(3): 271-275. doi: 10.3969/j.issn.1000-1158.2021.03.02
[18]	靳浩元, 刘军. 测量不确定度的评定方法及应用研究[J]. 计量科学与技术, 2021, 65(5): 124-131.
[19]	陈凌峰. 标准不确定度A类评定中极差法的深入讨论[J]. 计量学报, 2019, 40(2): 347-352.