【摘要】 受现场烟气取样位置及流场变化等因素影响，烟气氧量测量普遍存在多个信号测点数值偏差大、动态变化趋势不一致、可信性差的情况，为此提出一种基于卡尔曼滤波器的多点信号融合方法，得到准确性和可靠性更高的氧量信号。首先，利用增广状态卡尔曼滤波器，对空气预热器入口与烟囱入口氧量信号进行动态偏差估计与去噪，消除漏风引起的氧量偏差；其次，通过分布式卡尔曼滤波器，将初步融合信号与实时性最佳的炉膛出口氧量信号进一步融合，平衡动态响应与精度。经现场应用验证可知：在稳定工况下，当炉膛出口氧量方差为0.20、空预器入口氧量方差为0.25时，融合氧量的方差减小为0.011;在变负荷工况下，相较于烟囱入口氧量，融合氧量的等效惯性时间缩短了约200 s。融合后的氧量信号具有准确度高、动态响应速度快的优势，可提升氧量信号的可靠性。更多还原

【Abstract】 The oxygen content measurement generally suffers from large numerical deviations at multiple signal points, inconsistent dynamic change trends, and poor credibility due to factors such as the on-site flue gas sampling position and flow field changes. Therefore, a multi-point signal fusion method based on Kalman filter is proposed to obtain oxygen content signals with higher accuracy and reliability. Firstly, the augmented state Kalman filter is used to estimate the dynamic deviation and denoise the oxygen content signals at the air preheater inlet and chimney inlet to eliminate the oxygen content deviation caused by air leakage; secondly, the distributed Kalman filter is used to further fuse the preliminary fusion signal with the furnace outlet oxygen content signal, which has the best real-time performance, to balance the dynamic response and accuracy. Field application verification shows that the fused oxygen content signal has the advantages of high accuracy and fast dynamic response speed. Under stable operating conditions, the variance of the fused oxygen signal is reduced to 0.011 from 0.20 at the furnace outlet and 0.25 at the air preheater inlet. During load-changing conditions, the equivalent inertial time of the fused oxygen signal relative to the chimney inlet signal is shortened by approximately 200 s, which significantly enhances the reliability of the oxygen content signal.更多还原