基于ANN的泡沫金属阻隔爆效果预测研究
发布时间：2018-08-31 18:09
【摘要】：泡沫金属是目前能够同时抑制瓦斯爆炸火焰波和压力波的新型阻隔爆材料,影响其阻隔爆效果的因素众多,而测试实验过程费用高,周期长。为减少实验周期的材料损耗,优化泡沫金属参数组合,提高实验效率,采用人工神经网络方法对不同参数组合的泡沫金属阻隔效果进行了预测研究。结果表明:BP神经网络适用泡沫金属阻隔爆效果预测。当BP网络采用10个神经元,传递函数选择"logsig"、"purelin",网络达到最优,预测压力和温度的最大衰减率平均误差分别为13%和4%。研究表明人工神经网络可以用于泡沫金属阻隔爆效果的预测。
[Abstract]:Foam metal is a new type of flame barrier material which can simultaneously suppress the flame wave and pressure wave of gas explosion at present. There are many factors that affect the flame wave and pressure wave of gas explosion, but the cost of test process is high and the period is long. In order to reduce the material loss in the experiment cycle, optimize the combination of foam metal parameters and improve the experimental efficiency, the artificial neural network method was used to predict the barrier effect of different parameter combinations of foam metal. The results show that the 1: BP neural network is suitable for predicting the effect of foam metal explosion barrier. When the BP network adopts 10 neurons, the transfer function selects "logsig" and "purelin", and the network reaches the optimum. The average error of the maximum attenuation rate of predicting pressure and temperature is 13% and 4%, respectively. The results show that the artificial neural network can be used to predict the effect of foam metal explosion barrier.
【作者单位】： 黑龙江科技大学安全工程学院;中国矿业大学安全工程学院;黑龙江科技大学理学院;黑龙江龙煤鹤岗矿业有限责任公司俊德煤矿;
【基金】：黑龙江省自然科学青年基金资助项目(QC2015054) 国家安全监督管理总局2016重大事故防治关键技术科技项目(heilongjiang-0001-2016AQ)
【分类号】：TD712.7
本文编号：2215756
[Abstract]:Foam metal is a new type of flame barrier material which can simultaneously suppress the flame wave and pressure wave of gas explosion at present. There are many factors that affect the flame wave and pressure wave of gas explosion, but the cost of test process is high and the period is long. In order to reduce the material loss in the experiment cycle, optimize the combination of foam metal parameters and improve the experimental efficiency, the artificial neural network method was used to predict the barrier effect of different parameter combinations of foam metal. The results show that the 1: BP neural network is suitable for predicting the effect of foam metal explosion barrier. When the BP network adopts 10 neurons, the transfer function selects "logsig" and "purelin", and the network reaches the optimum. The average error of the maximum attenuation rate of predicting pressure and temperature is 13% and 4%, respectively. The results show that the artificial neural network can be used to predict the effect of foam metal explosion barrier.
【作者单位】： 黑龙江科技大学安全工程学院;中国矿业大学安全工程学院;黑龙江科技大学理学院;黑龙江龙煤鹤岗矿业有限责任公司俊德煤矿;
【基金】：黑龙江省自然科学青年基金资助项目(QC2015054) 国家安全监督管理总局2016重大事故防治关键技术科技项目(heilongjiang-0001-2016AQ)
【分类号】：TD712.7
【相似文献】
相关期刊论文 前3条
1 郭荣坤，王忠东;ANN测井相分析方法研究[J];地球物理学进展;1996年02期
2 王德润,谢广祥,孟祥瑞;基于ANN的综放工作面产量与工效预测[J];辽宁工程技术大学学报;2003年06期
3 胡敬朋;杨公训;;基于ANN的煤岩破坏电磁辐射预测研究[J];金属矿山;2009年09期
，本文编号：2215756
本文链接：https://www.wllwen.com/kejilunwen/anquangongcheng/2215756.html
