不完全量测下估计误差协方差与传感器精度上–下确-控制理论与应用.pdfVIP

30 4 Vol. 30 No. 4 2013 4 Control Theory Applications Apr. 2013 DOI: 10.7641/CTA.2013.20432 – , , , (, 475004) : , . – . . , – , . , , , , , . : ; ; ; : TP391 : A Infimum-supremum of sensor accuracy versus estimation error covariance in intermittent observations HU Zhen-tao, LI Song , LIU Xian-xing, LI Jie (Institute of Image Processing and Pattern Recognition, Henan University, Kaifeng Henan 475004, China) Abstract: In estimating system states in intermittent observations, the steady-state error covariance depends on the precision of sensors. We propose a novel algorithm for determining the required infimum-supremum of sensor accuracy based on the given estimation error covariance in intermittent observations. Starting from the estimation error covariance expressions of the steady-state Kalman filter; this algorithm determines the sensor detection probability and the admissible upper bound of the measurement noise covariance by using the linear matrix inequality. According to these results, the infimum requirement for sensor accuracies may be adopted under the given the estimation error covariance, thus relax- ing the requirements of sampling frequency and reducing the costs of measuring equipment as long as other engineering requirements are satisfied. Key words: state estimation; sensor accuracy; intermittent observations; covariance constraint 1 (Introduction) [3]