铁道科学与工程学报2024,Vol.21Issue(1) :342-353.DOI:10.19713/j.cnki.43-1423/u.T20230332

基于AMCPSO优化Kriging插值的温度补偿方法研究

Temperature compensation method based on AMCPSO Optimized Kriging interpolation

张森 王大志 黄晨涛 陈相吉 郑晓虎 刘梦哲
铁道科学与工程学报2024,Vol.21Issue(1) :342-353.DOI:10.19713/j.cnki.43-1423/u.T20230332

基于AMCPSO优化Kriging插值的温度补偿方法研究

Temperature compensation method based on AMCPSO Optimized Kriging interpolation

张森 1王大志 1黄晨涛 2陈相吉 1郑晓虎 1刘梦哲1
扫码查看

作者信息

  • 1. 大连理工大学 机械工程学院,辽宁 大连 116024
  • 2. 西安铁路信号有限责任公司,陕西 西安 710100
  • 折叠

摘要

为了降低温度变化对转换力传感器测量精度的影响,提出一种自适应变异混沌粒子群算法(AMCPSO)优化Kriging插值的温度补偿算法(AMCPSO-Kriging).研发转换力传感器,分析温度对传感器输出的影响,建立温度补偿标定实验平台,通过标定实验获得建立温度补偿模型所需要的样本集,采用数据稀疏化方法对样本数据进行优化.通过Kriging插值构建了温度补偿模型,利用AMCPSO算法以交叉验证方式下模型预测产生的均方根误差和作为适应度函数,对Kriging插值中的范围参数θ和平滑度参数pk进行寻优求解,得到性能最佳的温度补偿模型.基于AMCPSO-Kriging温度补偿模型对转换力传感器的测量效果进行实验验证,与标准力传感器进行对比.实验结果表明:对样本数据进行稀疏化处理,算法平均运行时间从1 076 s减少到6 s,提高了温度补偿算法的运行效率.在-20~70℃温度范围内,经过AMCPSO算法优化的Kriging模型有效提高了转换力传感器的测量精度,相比于未经AMCPSO算法优化的Kriging插值,转换力传感器测量的平均满量程误差从1.2%FS降低到0.6%FS.通过现场实验验证温度补偿的效果,转换力传感器测量的绝对误差在70 N以内,最大满量程误差为2.3%FS.所提出的温度补偿方法有效消除了温度对传感器测量精度的影响,满足铁路工况使用要求,对转换力传感器在铁路上实际运用具有重要价值.

Abstract

In order to reduce the influence of temperature variation on the measurement accuracy of the conversion force sensor,a temperature compensation algorithm(AMCPSO-Kriging)with an adaptive variational chaotic particle swarm algorithm(AMCPSO)optimized for Kriging interpolation was proposed.The conversion force sensor was developed,the effect of temperature on the sensor output was analyzed,and a temperature compensation calibration experimental platform was established.The sample set required to build the temperature compensation model was obtained through the calibration experiment,and the sample data was optimized by the data sparseness method.The temperature compensation model was constructed by Kriging interpolation.The AMCPSO algorithm was employed to search for the optimal solution for the range parameter θ and the smoothness parameter pk in Kriging interpolation by using the root-mean-square error sum of the model prediction as the fitness function under the cross-validation method to obtain the temperature compensation model with the best performance.Based on the AMCPSO-Kriging temperature compensation model,the measurement effect of the converted force sensor was experimentally verified and compared with the standard force sensor.It is found that the average running time of the algorithm is decreased from 1 076 s to 6 s by sparsifying of the sample data,and thus improving the operating efficiency of the temperature compensation algorithm.From-20℃to 70℃,the Kriging model optimized by AMCPSO algorithm effectively improved the measurement accuracy of the conversion force sensor.Meanwhile,the average full-scale error of the conversion force sensor measurement is decreased from 1.2%FS to 0.6%FS compared with the Kriging interpolation without AMCPSO algorithm.The effect of temperature compensation is proved through the field experiments.The absolute error of the conversion force sensor is within 70 N,while the maximum full range error is found to be 2.3%FS.The proposed temperature compensation method effectively eliminates the influence of temperature on the measurement accuracy of the sensor and meets the requirements of railroad working conditions,which is of great value for the practical application of the conversion force sensor on the railroad.

关键词

转换力传感器/温度补偿/标定实验/Kriging插值/自适应变异混沌粒子群优化算法

Key words

conversion force sensor/temperature compensation/calibration experiment/Kriging interpolation/adaptive mutation chaotic particle swarm optimization(AMCPSO)

引用本文复制引用

基金项目

国家自然科学基金资助项目(51975104)

国家自然科学基金资助项目(62074138)

出版年

2024
铁道科学与工程学报
中南大学 中国铁道学会

铁道科学与工程学报

CSTPCDCSCD北大核心EI
影响因子:0.837
ISSN:1672-7029
参考文献量9
段落导航相关论文