基于改进信度函数分配方法的煤种判别技术

针对电厂煤质不稳定的问题,提出基于D-S证据理论进行煤种判别的方法.提出一种根据以往的煤质分析数据建立可更新的典型样本库,然后根据典型样本库获得信度密度函数的方法.利用入炉煤元素分析结果和低位发热量作为证据,结合典型样本库计算出证据在各目标模式下的信度函数分配及不确定度,再利用D-S联合规则将数据进行融合,对目标煤种做...

Diagnosis technology of coal type based on improved belief function distribution method

For instability problem in coal properties of power plants,a method based on the D-S evidence theory to identify the coal type was proposed.An updated typical sample database based on the former analysis data of coal was established,then a new method based on the database was proposed to obtain the belief density function.The results of elemental analysis and net calorific value of burning coal in utility boiler can be used to train the evidence.Combined with the typical sampled database,the belief function assignment and uncertainty of the objective modes can be computed,and then the fusion belief function assignment was gotten by using D-S rule.Finally,the target coal can be determined.The method reduces the subjectivity of the judging process of coal evidently,and it can be used to monitor and control the varied coal in a power plant.