徐安桃,李锡栋,周慧,乔渊博,吴正日.基于灰色补偿BP神经网络优化组合模型的车辆装备冷却系腐蚀预测[J].装备环境工程,2018,15(11):123-128. XU An-tao,LI Xi-dong,ZHOU Hui,QIAO Yuan-bo,WU Zheng-ri.Corrosion Prediction of Vehicle Equipment Cooling System Based on Grey Compensation BP Neural Network Combined Model[J].Equipment Environmental Engineering,2018,15(11):123-128. |
基于灰色补偿BP神经网络优化组合模型的车辆装备冷却系腐蚀预测 |
Corrosion Prediction of Vehicle Equipment Cooling System Based on Grey Compensation BP Neural Network Combined Model |
投稿时间:2018-09-12 修订日期:2018-11-25 |
DOI:10.7643/ issn.1672-9242.2018.11.023 |
中文关键词: 车辆装备冷却系 MUGM(1,1,λ) 遗传算法 BP神经网络 |
英文关键词:vehicle equipment cooling system MUGM (1,1,λ) genetic algorithm BP neural network |
基金项目: |
|
Author | Institution |
XU An-tao | a. Projection Equipment Support Department, Army Military Transportation University, Tianjin 300161, China |
LI Xi-dong | b. Postgraduate Training Brigade, Fifth Team of Cadets, Army Military Transportation University, Tianjin 300161, China |
ZHOU Hui | b. Postgraduate Training Brigade, Fifth Team of Cadets, Army Military Transportation University, Tianjin 300161, China |
QIAO Yuan-bo | b. Postgraduate Training Brigade, Fifth Team of Cadets, Army Military Transportation University, Tianjin 300161, China |
WU Zheng-ri | b. Postgraduate Training Brigade, Fifth Team of Cadets, Army Military Transportation University, Tianjin 300161, China |
|
摘要点击次数: |
全文下载次数: |
中文摘要: |
目的 避免由于腐蚀破坏车辆装备冷却系,使车辆不能维持良好的工作状态,并缩短装备的使用寿命,建立一个准确、高效的预测模型,以实现对车辆装备冷却系腐蚀预测。方法 在传统GM(1,1)模型基础上,结合背景值构造优化与新陈代谢思想,建立一种新陈代谢加权不等时距模型MUGM(1,1,λ)。此外,还引入遗传算法优化BP神经网络模型对MUGM(1,1,λ)模型进行残差修正,建立灰色补偿BP神经网络优化组合模型。结果 基于优化组合模型对冷却系用铸铁材料腐蚀预测的平均误差为0.43%,模型精度为一级,预测精度高。结论 所建立的灰色补偿BP神经网络优化组合模型对于车辆装备冷却系金属腐蚀预测具有可行性。 |
英文摘要: |
Objective To avoid damage to the vehicle equipment cooling system due to corrosion, which disenable the vehicle to maintain a good working condition and shorten the service life of the equipment, and establish an accurate and efficient prediction model to predict corrosion of the vehicle equipment cooling system. Methods Based on the traditional GM (1,1) model, combined with background value construction optimization and metabolism, a metabolic weighted unequal time interval model MUGM(1,1,λ) was established. In addition, the genetic algorithm was used to optimize the BP neural network model to modify the residual of MUGM(1,1,λ) model, and the grey compensation BP neural network optimization combination model was established. Results Based on the optimized combination model, the average error of corrosion prediction for cast iron materials for cooling system was 0.43%; the model accuracy was level one; and the prediction accuracy was high. Conclusion The gray compensation BP neural network optimization combination model is feasible for the prediction of metal corrosion in vehicle cooling equipment. |
查看全文 查看/发表评论 下载PDF阅读器 |
关闭 |
|
|
|