矿井安全监测多源信息融合方法的研究

Study on Multi-source Information Fusion Method for Coal Mine Safety Monitoring

为了提高矿井安全监测的准确性,引入了一种基于改进证据理论的多源信息融合方法。该方法根据矿井四个不同空间域的特征信息,利用粒子群优化算法来构建矿井安全的初级识别模型,再由证据理论进行信息融合,实现对安全状况的判断。同时引入证据可信度,对初级识别模型的输出结果进行修正。试验结果表明,该方法在原始证据相一致以及出现高度冲突的情况下都具有较高的识别效果,实现了对矿井安全状态的有效监测。

To enhance the accurateness of coal mine safety monitoring,the multi-source information fusion method based on improved evidence theory is introduced. In accordance with the characteristic information in four different spatial domains,by adopting particle swarm optimization algorithm,the primary recognition model of coal mine safety is established; then information fusion is conducted by evidence theory,to judge the security conditions. In addition,the evidence credibility is introduced,for correcting the output result of the primary recognition model. The experimental results show that the method provides higher recognition effects in both conditions of original evidence is highly conflict or consistent,thus effective safety monitoring can be implemented for coal mine.