نوع مقاله : مقاله پژوهشی (کاربردی)
نویسندگان
1 استادیار، دپارتمان مهندسی مکانیک، آموزشکده شهید مفتح همدان، دانشگاه فنی و حرفه ای استان همدان، ایران .
2 استادیار، دپارتمان مهندسی مکانیک، دانشکده شهید چمران رشت، دانشگاه فنی و حرفهای استان گیلان، ایران.
3 عضو هیئت علمی، دپارتمان مهندسی مکانیک، آموزشکده امام خمینی بهشهر، دانشگاه فنی و حرفه ای استان مازندران، ایران .
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
One of the most important problems in control engineering is state estimation of a dynamical system based on measured data corrupted by noises. One of the most popular algorithms used for state estimation of a linear discrete time dynamical system is mixed Kalman/H-infinity filter. The performance of this filter is essentially depending on how exact statistics of noise characteristics are available. It is also not guaranteed that the process noise covariance matrix, and the measurement noise covariance matrix remain constant with time in a highly non-stationary noise condition. Thus, it is imperative to continuously tune the mixed Kalman/H-infinity accounting for the changing noise conditions in order to get good filter performance. This paper presents an algorithm of fuzzy based mixed Kalman/H-infinity filter for dynamically tuning the process noise and measurement noise covariance matrices. Fuzzy system in every step of the process using the difference between the actual position and speed of the balloon and the amount of observed data by the observer, an adaptive produce and this factor is used to update covariance matrix values. In this way, the Fuzzy Mixed Kalman/H-infinity filter can be more accurately estimated system state variables and therefore always mean square estimation error is minimized.
کلیدواژهها [English]