非线性间歇过程的迭代学习状态估计

在间歇过程的状态估计中,如何充分利用多批次重复特性信息是一个挑战。迭代学习卡尔曼滤波方法利用卡尔曼滤波沿时间方向估计相邻两批次之间的状态误差,并沿批次方向迭代更新当前状态估计,兼顾了时间和批次两维特性。但是,这种方法只适用于线性系统。针对非线性间歇过程,提出一种迭代学习拟线性卡尔曼滤波器(ILQKF)方法。ILQKF基于间歇过程的标称模型,将实际状态与标称状态之间的误差作为新状态,建立了与误差相关的线性化模型。然后,根据迭代学习卡尔曼滤波方法,对状态误差进行估计,而状态轨迹为误差轨迹与标称轨迹之和,从而估计出非线性间歇过程的状态。啤酒发酵过程的应用仿真验证了ILQKF方法的优越性。

Iterative learning state estimation for nonlinear batch process

In state estimation of batch process, it is a challenge to make full use of the multi-batch repetitive characteristic information. An iterative learning Kalman filter method uses Kalman filter to estimate the state errors between adjacent batches along the time direction, and iteratively updates the current state estimation along the batch direction, taking into account both time and batch two-dimensional characteristics. However, this method only applies to linear systems. An iterative learning quasilinear Kalman filter (ILQKF) method is proposed to estimate the state of nonlinear batch process. Based on the nominal model of batch process, ILQKF takes the error between the real state and the nominal state as a new state. Then a linearized model related to the error is established. The estimation of the error is obtained by referring to the method of the iterative learning Kalman filter. The state trajectory is equal to the sum of the error trajectory and the nominal trajectory, so as to estimate the state of the nonlinear batch process. The beer fermentation simulation is used to verify the performance of the ILQKF method.