滚动时域估计在三容系统中的应用

Application of moving horizon estimation in three tank system

为解决处理时域硬约束问题,基于预测控制的滚动优化原理,讨论了一种能用已知约束信息提高估计的合理与准确性的滚动时域估计(MHE:MovingHorizonEstimation)方法,并阐述了滚动时域估计的实用与理论特性。给出了三容系统的仿真结果并与Kalman估计进行了比较,结果表明,MHE方法在处理有约束状态估计方面比卡尔曼滤波方法更有效。

A MHE(Moving Horizon Estimation) approach that can make use of additional knowledge to achieve improvements in the estimation performance is presented. Some practical and theoretical properties of MHE are discussed. Simulation results for a three tank system and comparisons with the Kalman estimator are given,which indicate the more effectiveness of MHE than Kalman filter for the constrained linear system.