Sensor fusion algorithm

Kalman tace

Kalman tace yana hali.

Tushen algorithm shine saita saitin abubuwan ''imani'' ga kowane firikwensin. A kowane lokaci, za a yi amfani da bayanan firikwensin daga lokacin ƙarshe don ƙididdiga don inganta zato (ƙara kai), kuma za a yi la'akari da ingancin firikwensin. A cikin kwatanta tsakanin ƙimar da aka annabta da ƙimar ƙimar firikwensin, za a ƙididdige ƙima mai kyau da fitarwa.

Wannan yana nufin cewa idan na'urar firikwensin koyaushe yana ba da ƙima mai kyau da daidaito kuma ya fara gaya muku wani abu da ba zai yuwu ba, ƙimar amincin firikwensin zai ragu a cikin 'yan milli seconds har sai ya fara yin hankali.

Wannan ya fi sauƙaƙa matsakaici ko jefa ƙuri'a, saboda Kalman tace zai iya magance yanayin da yawancin firikwensin ba su da tsari na ɗan lokaci. Muddin mutum zai iya kiyaye kyakkyawan dalili, zai iya sanya mutum-mutumin cikin duhu.

Kalman tace aikace-aikace ne na ƙarin fahimtar sarkar Markov da kuma tunanin Bayesian, wanda shine tsarin lissafi wanda ke inganta hasashensu akai-akai ta hanyar amfani da shaida. Waɗannan kayan aikin kayan aiki ne da ake amfani da su don taimakawa kimiyya da kanta ta gwada ra'ayoyi (waɗanda kuma su ne tushen abin da muke kira "mahimmancin ƙididdiga").

Don haka, za a iya cewa ta hanyar waka cewa wasu na'urorin haɗakar bayanai suna bayyana ainihin kimiyya a gudun sau 1000 a cikin daƙiƙa guda.

An yi amfani da filtatan Kalman a tashoshin sararin samaniya na tsawon shekaru da yawa. Saboda microcontrollers na zamani na iya tafiyar da algorithm a ainihin lokacin, suna ƙara samun shahara a cikin injiniyoyi.