论文部分内容阅读
本文借助于统计理论,从误差方差分析角度对迭代滤波方法较推广卡尔曼滤波方法的优越性给出数学证明,为迭代滤波方法能够成功地运用于生化反应过程的状态估计提供理论依据;进而对几个典型的实际生化反应过程应用迭代滤波方法进行状态估计作了大量的数字仿真研究。结果表明,迭代滤波方法是完全能够成功地应用到生化反应过程中,解决其状态估计向题是行之有效的。对于非线性程度严重的生化反应过程,迭代滤波方法相比于推广卡尔曼滤波方法有显著改进。应用迭代滤波方法能够从带有强烈随机噪声的部分检测信息中获取与实际过程的真实状态向应存在良好一致性的全部状态估计,起到了在线观测器的作用为克服生化反应过程中许多变量难以在线观测这一困难提供了一条有效途径。
Based on the statistical theory, this paper gives a mathematical proof of the superiority of the iterative filtering method over the extended Kalman filter from the perspective of error variance analysis, and provides a theoretical basis for the state estimation of the iterative filtering method successfully applied to the biochemical reaction process. Several typical actual biochemical reaction process using iterative filtering method for state estimation made a lot of digital simulation. The results show that the iterative filtering method can be successfully applied to the biochemical reaction process, and it is effective to solve the problem of state estimation. For the non-linear biochemical reaction process, the iterative filtering method has significantly improved compared with the extended Kalman filtering method. The application of iterative filtering method can obtain the full state estimation from the partial detection information with strong random noise which is in good agreement with the true state of the actual process and plays the role of on-line observer. In order to overcome the difficulty of many variables in biochemical reaction process An effective way to observe this difficulty online is.