化工学报 ›› 2011, Vol. 62 ›› Issue (S1): 219-224.

• 能源和环境工程 • 上一篇    下一篇

基于最大熵原理的卧罐爆炸碎片数量概率分布

孙东亮,蒋军成,张明广,王志荣   

  1. 南京工业大学城市建设与安全工程学院,江苏省城市与工业安全重点实验室,江苏 南京 210009
  • 出版日期:2011-07-03 发布日期:2011-07-03

  • Online:2011-07-03 Published:2011-07-03

摘要:

分析了储罐爆炸碎片数量与罐形、事故类型、破坏机理的关联性,通过卧罐爆炸事故数据的汇总研究,基于较多的数据,通过统计分析获得了四类爆炸事故数量及碎片数量与分布比例。依据最大熵原理,并通过爆炸事故数据的分析获得可利用的已知信息为碎片数量均值及方差,由此建立四类爆炸事故产生的碎片数量的概率分布模型。结果表明,各爆炸类型生成的碎片数量区间大体上均为[1, 9],且各类事故中碎片数量14均占有较大比例,不同事故类型对应的碎片数量分布比例不同。利用建立的模型计算获得的碎片数量理论概率分布与实际数据的吻合性较好。碎片数量在统计区间[1, 9]内均服从离散指数分布。

关键词: 卧罐爆炸, 碎片数量, 最大熵原理, 概率分布模型

Abstract:

The relations between the fragments number from the vessel explosion, the tank shape, the type of the primary scenario and the fracture mechanics were analyzed.The data of the past explosions leading to the fragment projection were collected and researched, and then the number of events for the four explosion categories of the horizontal tank and the corresponding observed probabilities of the fragments number were obtained by statistics on the basis of the more available data.The available information including the average and variance of the fragments number was obtained by analyzing the accident data.The probability density functions for the fragments number from the four categories of explosions were developed by the maximum entropy principle.The results showed that all the intervals of the number of fragments from each type of accident were generally 1, 9, the fragments number 1, 2, 3 and 4 account for larger proportions in any event, and the fragments number and their observed probabilities were different for different accidental events.The theoretical histograms derived by the four models were compared to the accident frequencies.The fragments number followed the discrete exponential distributions within the interval 1, 9.

Key words: 卧罐爆炸, 碎片数量, 最大熵原理, 概率分布模型