摘要
模型材料是水工结构模型试验成功的关键,但其制备方法较为依赖经验、成本较高,且试验数据利用率较低。为此,采用单因素敏感性分析方法分析了胶凝砂砾石模型材料弹性模量的影响因子,并基于敏感性分析结果利用Matlab建立了BP神经网络预测模型。预测检验结果表明,水工结构模型材料的弹性模量对拌合水量和石膏量最敏感,其次是重晶石粉量,对铁粉量最不敏感;在低拌合水量下,对水泥量、石蜡量和机油量敏感性不高。可通过控制低敏感性材料的用量,制备较高精度弹性模量的模型材料。预测模型可靠度达R^(2)=0.939,满足对数据预测的精度要求,弥补了试验人员的经验不足。
In the hydraulic structure model test, the model material is the key to the test’s success. However, the preparation method is more dependent on experience, the cost is high, and the test data utilization rate is low. Therefore, the single factor sensitivity analysis method was used to analyze the elastic modulus’ s influencing factors of the cemented sand and gravel model material. Furthermore, based on the analysis results, a BP neural network prediction model was established. The results show that the elastic modulus of hydraulic structure model materials is the most sensitive to mixing water and gypsum, followed by barite powder, the least sensitive to iron powder. At low mixing water, the amount of paraffin wax and engine oil is not sensitive. By controlling the amount of low sensitive material, the model material with high precision elastic modulus can be prepared. The reliability of the prediction model is R^(2)=0.939, which meets the precision requirements of data prediction and makes up for the lack of experience of the test personnel.
作者
丁泽霖
高昱芃
张宏洋
王婧
DING Ze-lin;GAO Yu-peng;ZHANG Hong-yang;WANG Jing(School of Water Conservancy,North China University of Water Resources and Electric Power,Zhengzhou 450046,China)
出处
《水电能源科学》
北大核心
2022年第3期158-161,共4页
Water Resources and Power
基金
国家重点研发计划(2018YFC0406803)
国家自然科学基金青年基金项目(51709114)。
关键词
水工结构模型
模型材料
敏感性分析
影响因子
BP神经网络
弹性模量
hydraulic structure model
model material
sensitivity analysis
influencing factor
BP neural network
elastic modulus
