定义了混煤和锅炉目标煤的煤质偏差，给出以煤质偏差作为目标函数的掺配优化模型，通过调整某个成分的期望误差，可单独调节该成分与目标煤的接近程度；通过对模型约束条件设定，可限制低灰熔点煤的掺配份额，控制炉膛结焦，或提高采购量大的单煤的掺配比例。基于最小煤质偏差模型，针对某670 MW超临界压力直流锅炉进行掺配优化模拟，提出煤场存放的优化概念和方法。结果表明：该模型能够从复杂的单煤煤源中找到最接近目标煤质的掺配方案，优化后入炉煤质稳定，且接近目标煤质。

Based on definition of quality deviation,a novel procedure for coal blending optimization has been developed in this paper.In this procedure,the deviation of a composition of coal blends from that of target coal can be adjusted independently with the expected error of the composition.By setting the restriction conditions,the mass fractions of a coked coal can be reduced,or the blending ratio of large quantity of single coal can be increased.The minimum deviation procedure was applied to a 670 MW supercritical pressure oncethrough boiler in order to optimize the storage and distribution of various coal in coal yard.Results showed that minimum deviation blending scheme has been determined,and the quality of the blending coal was stable and extremely close to the target coal.

TK16

上海市科委项目（14110502400）