新型融合估计方法及其在成纸定量中的应用

New Fusion Algorithm and Its Application in Estimation of Paper Basis Weight

在工业生产中造纸机状态监测的工程实际中,使用多个同类传感器进行在线测量可以得到更为准确的状态估计.在这类特殊的多传感器系统中,本文通过矩阵运算消除相关估计方差,得到了最优分布式融合估计算法.与多传感器传统分布式次优融合算法相比,该算法的优点在于适用于含有多于三个多传感器的系统.由于不需要存贮和计算中间变量,在计算量方面比最优融合算法具有较大优势.应用结果表明算法具有良好的估计性能.

In practice, the state supervision of paper machine is generally obtained by the same kind of sensors, which can perform a more estimation result. For this special multisensor system, distributed optimal fusion algorithm is received by avoiding computing correlated estimation covariance based on the matrix operation. Compared with classical multisensor distributed-suboptimal algorithm and optimal fusion algorithm, this algorithm can adapt to the system with more than three sensors and has the advantages of the count capacity because it has no use for saving and computing the middle variable. Applications show the developed algorithm has the excellent estimation performance.