تخمين خطاي موقعيت سيستم ناوبري اينرسي از طريق تلفيق با سامانه تصويري

Inertial navigation position error estimation, using vision system

ﻣﺴﺌﻠﻪ اﺻﻠﻲ ﻣﻮرد ﻣﻄﺎﻟﻌﻪ در اﻳﻦ ﻣﻘﺎﻟﻪ، ﺗﺨﻤﻴﻦ ﺧﻄﺎي ﻣﻮﻗﻌﻴﺖ و وﺿﻌﻴﺖ ﺳﻴﺴﺘﻢ ﻧﺎوﺑﺮي اﻳﻨﺮﺳﻲ از ﻃﺮﻳﻖ ﺗﻠﻔﻴﻖ ﺑﺎ دادهﻫﺎي ﺳﺎﻣﺎﻧﻪي ﺑﻴﻨﺎﻳﻲ اﺳﺖ. ﺑﺴﺘﺮ ﻣﻮرد ﻣﻄﺎﻟﻌﻪ، ﻳﻚ ﻓﻀﺎﭘﻴﻤﺎي ﺑﺎزﮔﺸﺘﻲ اﺳﺖ ﻛﻪ ﺑﺎﻳﺪ ﻣﻮﻗﻌﻴﺖ ﺧﻮد را ﻧﺴﺒﺖ ﺑﻪ ﻳﻚ ﻧﻘﻄﻪ ﻓﺮود از ﭘﻴﺶ ﺗﻌﻴﻴﻦ ﺷﺪه، ﺑﻄﻮر دﻗﻴﻖ اﻧﺪازهﮔﻴﺮي ﻛﻨﺪ. ﻓﺮض ﺷﺪه اﺳﺖ ﻛﻪ ﻓﻀﺎﭘﻴﻤﺎ از ﻳﻚ ﺳﻴﺴﺘﻢ ﻧﺎوﺑﺮي ﻣﺎﻫﻮارهاي ﻛﻤﻜﻲ ﺑﻬﺮه ﻣﻲﮔﻴﺮد. ﺑﻨﺎﺑﺮاﻳﻦ در ﻣﻮاﻗﻌﻲ ﻛﻪ ﺳﻴﮕﻨﺎل ﻣﺎﻫﻮارهﻫﺎ ﻗﻄﻊ ﺷﻮﻧﺪ ﻳﺎ در ﺣﺎﻟﺘﻲ ﻛﻪ ﻓﺮود ﺑﺮ ﻳﻚ ﺳﻜﻮي درﻳﺎﻳﻲ ﻣﺘﺤﺮك ﻣﻮردﻧﻈﺮ ﺑﺎﺷﺪ، دادهﻫﺎي ﺳﻴﺴﺘﻢ ﻧﺎوﺑﺮي ﺗﺼﻮﻳﺮي، ﺟﺎﻳﮕﺰﻳﻦ اﻃﻼﻋﺎت ﺳﻴﺴﺘﻢ ﻧﺎوﺑﺮي ﻣﺎﻫﻮارهاي ﺷﺪه و ﺑﺎﻋﺚ ﺑﻬﺒﻮد دﻗﺖ ﺳﻴﺴﺘﻢ ﻧﺎوﺑﺮي ﻓﻀﺎﭘﻴﻤﺎ ﻣﻲﺷﻮﻧﺪ. ﺑﺮاي ﺗﻠﻔﻴﻖ اﻃﻼﻋﺎت ﺳﻴﺴﺘﻢ ﻧﺎوﺑﺮي اﻳﻨﺮﺳﻲ و دادهﻫﺎي ﺳﻴﺴﺘﻢ ﺗﺼﻮﻳﺮي از ﻓﻴﻠﺘﺮﻛﺎﻟﻤﻦ ﺗﻮﺳﻌﻪﻳﺎﻓﺘﻪ اﺳﺘﻔﺎده ﺷﺪه اﺳﺖ. ﺿﻤﻦ آن ﻛﻪ دادهﻫﺎي ﺧﺮوﺟﻲ ﺳﻴﺴﺘﻢ ﺗﺼﻮﻳﺮي ﺑﻪ ﻣﻨﻈﻮر اﺳﺘﻔﺎده در ﻣﻌﺎدﻻت اﻧﺪازهﮔﻴﺮي ﻓﻴﻠﺘﺮ ﻛﺎﻟﻤﻦ، اﺑﺘﺪا ﺑﻪ وﺳﻴﻠﻪ ﻓﻴﻠﺘﺮ ﺣﺪاﻗﻞ ﻣﺮﺑﻌﺎت ﺑﺎزﮔﺸﺘﻲ ﻣﻮرد ﭘﺮدازش ﻗﺮار ﻣﻲﮔﻴﺮﻧﺪ. در روش ﻣﺘﺪاول، از ﻣﺪل و ﺷﺒﻴﻪﺳﺎزي ﺳﻴﺴﺘﻢﻫﺎ و ﺳﺨﺖاﻓﺰارﻫﺎي ﻣﺨﺘﻠﻒ، ﺑﺮاي ﺗﺼﺪﻳﻖ و ﺣﺴﺎﺳﻴﺖﺳﻨﺠﻲ دﻗﺖ روش ﻧﺴﺒﺖ ﺑﻪ ﭘﺎراﻣﺘﺮﻫﺎي ﻫﺮ ﻗﺴﻤﺖ اﺳﺘﻔﺎده ﻣﻲﻛﻨﻨﺪ. ﺑﺎ اﻳﻦ روﻳﻜﺮد، ﻣﺪل- ﻫﺎي ﻣﻮرد ﺑﺤﺚ در اﻳﻦ ﻣﻘﺎﻟﻪ، ﻣﺮﺑﻮط ﺑﻪ ﺷﺒﻴﻪﺳﺎزي ﭘﺮواز، ﺗﺼﻮﻳﺮ دورﺑﻴﻦ، ﻣﻌﺎدﻻت ﻧﺎوﺑﺮي اﻳﻨﺮﺳﻲ، دﻳﻨﺎﻣﻴﻚ ﺧﻄﺎي ﺳﻴﺴﺘﻢ ﻧﺎوﺑﺮي اﻳﻨﺮﺳﻲ، ﻧﺎوﺑﺮي ﺗﺼﻮﻳﺮي و ﻓﻴﻠﺘﺮ ﻛﺎﻟﻤﻦ ﺗﻮﺳﻌﻪﻳﺎﻓﺘﻪ ﻫﺴﺘﻨﺪ. رواﺑﻂ ﻫﺮ ﺑﺨﺶ آورده ﺷﺪه و ﺑﺮاﺳﺎس ﻧﺘﺎﻳﺞ ﺷﺒﻴﻪﺳﺎزي ﻧﺮماﻓﺰاري، ﻛﺎراﻳﻲ روش ﭘﻴﺸﻨﻬﺎدي ﻧﺸﺎن داده ﺷﺪه اﺳﺖ.

The main task of the study is to estimate the position error in an inertial navigation system by integrating it with the visual system. The case study is a spacecraft that must accurately measure its position relative to a predetermined landing point. The spacecraft is assumed to be augmented GNSS navigation. Therefore, when satellite signals are dropped out or when landing on a moving marine platform, the data of the vision navigation system replaces the information of the satellite navigation system and improves the accuracy of the spacecraft navigation system. An Extended Kalman filter has been used to integrate inertial and vision navigation system information. In addition, the output data of the vision system, in order to be used in the Kalman filter measurement equations, is first processed by the recursive least square filter. The relevant relations are given and based on the results of software simulation, the efficiency of the proposed method is shown.