模糊神经网络在矿震预测中的应用
|
摘要
矿震同天然地震一样会给矿山生产及人身安全等带来重大灾难。也是目前尚不能准确预测的矿山灾害现象之一。根据现有的研究成果可知,矿震是一个多输入、多干扰、单输出的复杂系统。由于干扰项的存在,使利用建模、神经网络等手段对系统进行预测时会导致很大误差。模糊神经网络系统在建立对象输入、输出关系时与传统数学方法不同。即可以建立在无模型基础上,并利用其较强的学习训练特性,自动获取对象的输入、输出关系表达;可以将专家的评价语言作为系统的干扰项引入。这在一定程度上缓解了人为因素对预测结果的影响,且平滑了观测数据的随机性。文章利用改进的模糊神经网络及抚顺老虎台矿的矿震资料,对矿震最大震级的预测方法进行了探索。初步探讨了改进的模糊神经网络在矿震预测中的应用。得出在运用模糊神经网络进行预测时,为减小预测误差,应综合多种因素并提高专家评判语言的精确度的结论。指出在建立矿震系统预测模型时,利用干扰项将人为因素引入系统是必需的。通过实际应用证明其可行性。
The mining earthquake makes a lot of damage to people and mining production just as natural earthquake. It is one of mining hazards that could not be predicted accurately. According to research findings, the mining earthquake is a complex system that characterized a lot of causes, many conditions and single sequent. The lots of influence factors often make a great error when the system is predicted by quantitative models. Fuzzy neural network is different from the traditional mathematic method. It is established on the base without model and may get the relationship automatically between importation(factors) and exportation(results) by itself. And it may introduce experts' evaluating language as disturbance factors of the system, so it can weaken the influence of man-induced factors and smooth the randomness of observational data. This paper discuses the application of the improved fuzzy neural network system in mine earthquake prediction by using improved fuzzy neural network system and the date of the mine tremor of Laohutai in Fushun Coal Mine. All kinds of factors should be inducted as importation in the system and the precision of estimation from authority should be improved for reducing errors during using fuzzy neural network system to forecast the mine earthquake. It is necessary to introduce artificial factors into the system while establishing a model of prediction for mine earthquake. And the feasibility of prediction of mine earthquake using fuzzy neural network system has been proved in practice.
The mining earthquake makes a lot of damage to people and mining production just as natural earthquake. It is one of mining hazards that could not be predicted accurately. According to research findings, the mining earthquake is a complex system that characterized a lot of causes, many conditions and single sequent. The lots of influence factors often make a great error when the system is predicted by quantitative models. Fuzzy neural network is different from the traditional mathematic method. It is established on the base without model and may get the relationship automatically between importation(factors) and exportation(results) by itself. And it may introduce experts' evaluating language as disturbance factors of the system, so it can weaken the influence of man-induced factors and smooth the randomness of observational data. This paper discuses the application of the improved fuzzy neural network system in mine earthquake prediction by using improved fuzzy neural network system and the date of the mine tremor of Laohutai in Fushun Coal Mine. All kinds of factors should be inducted as importation in the system and the precision of estimation from authority should be improved for reducing errors during using fuzzy neural network system to forecast the mine earthquake. It is necessary to introduce artificial factors into the system while establishing a model of prediction for mine earthquake. And the feasibility of prediction of mine earthquake using fuzzy neural network system has been proved in practice.
引文
[1] 惠乃玲,刘耀权,杨明皓,等.抚顺老虎台煤矿矿震震源机制的研究[J].地震地磁观测与研究,1998,19(1):39-57.
[2] 孙文福,周元夫,焦明若,等.关于抚顺矿震活动趋势分析[J].高原地震,2002,14(2):45-51.
[3] 李铁,冀林旺,左艳,等.预测较强矿震的地震学方法探讨[J].东北地震研究,2003,19(1):53-59.
[4] 王士同.模糊系统、模糊神经网络及应用程序设计[M].上海:上海科学技术出版社,1998.
[5] 王红旗,余发山,郝加臣.汽包水位的模糊神经网络预测模型研究[J].自动化控制与应用,2003,22(2):1-3.
[6] 姜德义,蒋再文,章永武,等.模糊神经网络预测公路隧道揭煤突出危险性[J].重庆大学学报,2000,(51):200-202.
[7] 尹宏宾,徐建闽,刘列根.利用模糊神经网络预测道路交通事故损失[J].公路交通科技,1999,16(3):57-62.
[2] 孙文福,周元夫,焦明若,等.关于抚顺矿震活动趋势分析[J].高原地震,2002,14(2):45-51.
[3] 李铁,冀林旺,左艳,等.预测较强矿震的地震学方法探讨[J].东北地震研究,2003,19(1):53-59.
[4] 王士同.模糊系统、模糊神经网络及应用程序设计[M].上海:上海科学技术出版社,1998.
[5] 王红旗,余发山,郝加臣.汽包水位的模糊神经网络预测模型研究[J].自动化控制与应用,2003,22(2):1-3.
[6] 姜德义,蒋再文,章永武,等.模糊神经网络预测公路隧道揭煤突出危险性[J].重庆大学学报,2000,(51):200-202.
[7] 尹宏宾,徐建闽,刘列根.利用模糊神经网络预测道路交通事故损失[J].公路交通科技,1999,16(3):57-62.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心