یک روش ترکیبی هوشمند جدید مبتنی بر فیلتر کالمن و شبکه عصبی رگرسیون تعمیم یافته برای تلفیق سیستم ناوبری اینرسی ارزان‌قیمت و سیستم‌ ناوبری ماهواره‌ای جهانی

ﺑﺎ ﺗﻮﺟﻪ ﺑﻪ ﻋﺪم ﮐﺎرآﯾﯽ ﻣﻨﺎﺳﺐ روشﻫﺎی ﻣﺒﺘﻨﯽ ﺑﺮ ﻓﯿﻠﺘﺮ ﮐﺎﻟﻤﻦ ﺑﺮای ﺗﻠﻔﯿﻖ دادهﻫﺎی ﺳﯿﺴﺘﻢ ﻧﺎوﺑﺮی اﯾﻨﺮﺳﯽ ارزانﻗﯿﻤﺖ و ﺳﯿﺴﺘﻢﻫﺎی ﻧﺎوﺑﺮی ﻣﺎﻫﻮارهای ﺟﻬﺎﻧﯽ در زﻣﺎن ﻗﻄﻊ ﺷﺪن ﺳﯿﮕﻨﺎلﻫﺎی ﻣﺎﻫﻮارهای، اﺳﺘﻔﺎده از ﺗﮑﻨﯿﮏﻫﺎی ﻫﻮش ﻣﺼﻨﻮﻋﯽ در ﻣﻌﻤﺎری ﺗﻠﻔﯿﻖ ﻣﺮﺳﻮم ﺷﺪه اﺳﺖ. از اﯾﻦ رو در اﯾﻦ ﻣﻘﺎﻟﻪ ﺿﻤﻦ اراﺋﻪی ﯾﮏ ﻣﻌﻤﺎری ﺗﺮﮐﯿﺒﯽ ﻣﺆﺛﺮ، از ﺷﺒﮑﻪ ﻋﺼﺒﯽ رﮔﺮﺳﯿﻮن ﺗﻌﻤﯿﻢﯾﺎﻓﺘﻪ ﺑﺮای ﭘﯿﺶﺑﯿﻨﯽ ﻣﺸﺎﻫﺪات ﻣﻮرد ﻧﯿﺎز ﻓﯿﻠﺘﺮ ﮐﺎﻟﻤﻦ در ﺷﺮاﯾﻂ ﻗﻄﻊ ﺷﺪن ﻃﻮﻻﻧﯽ ﻣﺪت ﺳﯿﮕﻨﺎلﻫﺎی ﻣﺎﻫﻮارهای اﺳﺘﻔﺎده ﺷﺪه اﺳﺖ. در ﻣﺪل ﭘﯿﺸﻨﻬﺎدی، ﺑﺮای آﻣﻮزش ﺷﺒﮑﻪ ﻋﺼﺒﯽ، ﺳﺮﻋﺖﻫﺎ و ﻣﻮﻗﻌﯿﺖﻫﺎی ﺳﯿﺴﺘﻢ اﯾﻨﺮﺳﯽ ﺑﻪﻋﻨﻮان ورودیﻫﺎ و ﻧﯿﺰ ﺳﺮﻋﺖﻫﺎ و ﻣﻮﻗﻌﯿﺖﻫﺎی ﺳﯿﺴﺘﻢ ﻣﻮﻗﻌﯿﺖﯾﺎب ﺟﻬﺎﻧﯽ ﺑﻪﻋﻨﻮان ﺧﺮوﺟﯽﻫﺎی ﺷﺒﮑﻪ در ﻧﻈﺮ ﮔﺮﻓﺘﻪ ﺷﺪهاﻧﺪ. اﯾﻦ روﯾﮑﺮد در ﻋﯿﻦ ﮐﺎرﺑﺮدی و ﻋﻤﻠﯿﺎﺗﯽ ﺑﻮدن، ﺳﺒﺐ ﮐﺎﻫﺶ زﻣﺎن ﻣﺤﺎﺳﺒﺎﺗﯽ و اﻓﺰاﯾﺶ دﻗﺖ و ﺳﺮﻋﺖ آﻣﻮزش و ﺗﺨﻤﯿﻦ ﺷﺒﮑﻪ ﺷﺪه اﺳﺖ. ﻧﺘﺎﯾﺞ ﺷﺒﯿﻪﺳﺎزیﻫﺎ ﻧﺸﺎن ﻣﯽدﻫﻨﺪ ﮐﻪ ﺑﻪدﻟﯿﻞ ﺳﺎﺧﺘﺎر ﺳﺎده و در ﻋﯿﻦ ﺣﺎل ﻣﻘﺎوم ﻣﻌﻤﺎری ﭘﯿﺸﻨﻬﺎدی و اﻟﺒﺘﻪ اﻧﺘﺨﺎب ﯾﮏ ﺷﺒﮑﻪ ﻋﺼﺒﯽ ﮐﺎرآﻣﺪ ﭼﻨﺪ ورودی- ﭼﻨﺪ ﺧﺮوﺟﯽ ﺑﺎ ﻗﺎﺑﻠﯿﺖ ﮐﺸﻒ ارﺗﺒﺎط ﻣﺆﺛﺮ ﻣﯿﺎن ورودیﻫﺎ و ﺧﺮوﺟﯽﻫﺎی ﺗﻌﯿﯿﻦ ﺷﺪه و ﺑﻪ ﺗﺒﻊ آن اﺻﻼح ﻣﻨﺎﺳﺐ ﺧﻄﺎﻫﺎی ﻣﺮﺑﻮط ﺑﻪ ﺳﺮﻋﺖﻫﺎ و ﻣﻮﻗﻌﯿﺖﻫﺎی ﺳﯿﺴﺘﻢ ﻧﺎوﺑﺮی اﯾﻨﺮﺳﯽ، ﻣﯽﺗﻮان از آن ﺑﺮای ﻧﺎوﺑﺮی زﻣﺎن واﻗﻌﯽ، ﺧﻮداﺗﮑﺎ، ﺑﺎ ﻗﺎﺑﻠﯿﺖ اﻃﻤﯿﻨﺎن و دﻗﺖ ﺑﺎﻻ اﺳﺘﻔﺎده ﻧﻤﻮد.

A new intelligent hybrid method based on Kalman filter and GRNN for low-cost INS/GNSS integration

Due to the inefficiency of Kalman filterbased methods for combining lowcost inertial navigation system data and global satellite navigation systems when satellite signals are outage, the use of artificial intelligence techniques in integrated architecture has become a common issue. Therefore, in this paper, while presenting an effective hybrid architecture, the generalized regression neural network is used to predict the required observations of the Kalman filter at the event of longterm outage of satellite signals. In the proposed model, for training the neural network, the velocities and positions of the inertial system are considered as inputs and also the velocities and positions of the global positioning system are considered as network outputs. This approach, while being practical and operational, has reduced computational time and increased the accuracy and speed of training and network estimation. The simulation results show that due to the simple yet robust structure of the proposed architecture and of course the selection of an efficient multiinputmultioutput neural network with the ability to detect the effective relationship between inputs and specified outputs and consequently correct errors related to speeds and situations, inertial navigation system can be used for realtime navigation, selfreliant, with high reliability and accuracy.