论文部分内容阅读
为解决煤矿粉尘监测中传感器测量指标种类单一、测量数据量大等问题,提出了以粉尘浓度和粉尘粒度为监测对象,运用数据融合技术处理传感器信息的新方法。在建立数据融合两级结构模型的基础上,先应用基于矩阵分析的融合算法对同质源数据进行数据级融合,再应用D-S证据理论对异质源数据进行决策级融合,最终实现传感器信息的整合优化。试验和应用结果表明,该方法在完善粉尘表征评价指标的同时,显著提高了传感器信息的准确度和可信度。
In order to solve the problems of single type of measurement index and large amount of measurement data in coal mine dust monitoring, a new method to process the sensor information by using data fusion technology is put forward, which is based on dust concentration and dust particle size. Based on the data fusion two-level structure model, the fusion algorithm based on matrix analysis is applied to the data fusion of homogenous source data, and then the DS evidence theory is used to fuse the heterogeneous source data at the decision-making level. Finally, the sensor information Integration and optimization. The test and application results show that this method improves the accuracy and reliability of sensor information significantly while improving the evaluation index of dust characterization.