【作者】 王传英；

【导师】 付华；

【作者基本信息】 辽宁工程技术大学 ， 控制理论与控制工程， 2006， 硕士

【摘要】 矿井火灾的防治一直是煤炭工业安全生产的重点问题。由于矿井下的综合环境比较复杂,各种影响因素较多,使用传统的单一预测方法不能可靠有效的预测煤层自然发火。本文采用信息融合技术,综合矿井内多源信息,实现煤矿自然发火的精确预测。信息融合是信息领域的新型理论和技术。信息融合对于多源信息的处理能力得到了广泛应用,本文针对它在煤层自然发火预测中的应用作了系统的研究,提出了多种人工智能相结合的融合方法。本文利用信息融合将煤矿井下检测的多种信号进行融合处理。采用模糊和遗传算法,融合过程中的推理由模糊评判函数完成,模糊评判函数的参数由遗传算法进行优化确定。该算法的优点是在信息源的可靠性、冗余性及互补性不确定的情况下能够以近似最优的方法对信息进行融合。通过信息融合,本文得出了矿井环境的自然发火状态的预测系统,通过测试验证了此算法预测的可靠性。更多还原

【Abstract】 Mine fire preventing and controlling always is the key question of coal industry safety in production. Traditional sole forecast method cannot forecast coal spontaneous combustion reliably Because of complex combined environment and all kinds of influence factors under mine. Information fusion technology is used in this article and realizes precise forecast of coal spontaneous combustion. Information fusion is a new theory and technology in the information field. It is applied on the multiple source information handling widely. In view of its application in the coal bed spontaneous combustion forecast, it is researched systematically, multiple artificial intelligence unified fusion method is proposed. This article adopts information fusion to process many kinds of signals fusion. Fuzzy and genetic algorithms are Uses. The fusion process is completed by fuzzy assessment function and the fuzzy parameters are optimized and confirmed by genetic algorithms. The algorithm merit is information fusion by the approximate superior method in the situation of uncertain information source reliability, redundant and complementarity. The mine spontaneous combustion forecast system is obtained through information fusion in this article, forecast reliability is validated through the algorithm test.更多还原