Drone autopilot with neural network
سال انتشار: 1401
نوع سند: مقاله کنفرانسی
زبان: انگلیسی
مشاهده: 1,571
فایل این مقاله در 14 صفحه با فرمت PDF قابل دریافت می باشد
- صدور گواهی نمایه سازی
- من نویسنده این مقاله هستم
این مقاله در بخشهای موضوعی زیر دسته بندی شده است:
استخراج به نرم افزارهای پژوهشی:
شناسه ملی سند علمی:
CARSE06_244
تاریخ نمایه سازی: 26 اردیبهشت 1401
چکیده مقاله:
UAVs have different uses, although navigation is done by human operators or independently. But flight relies on in-flight sensors, such as inertia units (IMUs) and Global Positioning Systems (GPS), to steer the drone independently in predefined directions. Such an approach typically integrates GPS location with a CCTV navigation control system using the IMU. IMU and GPS feedback allow the drone to move in certain directions and correct any drift. The purpose of this paper is to guide the drone by internal sensors such as inertia units (IMUs) and Global Positioning Systems (GPS) to steer the drone independently and intelligently in predefined directions. In this research, statistical community method, statistical sampling, sampling and cyberspace tools, validity and reliability study and diagnostic analysis with spatial analysis have been used and research has been performed on UAV autopilot with neural network. The fundamental and applied method is effective in the present study. Findings include system description, related work, methodology, control, flight enhancement, evaluation criteria, data set, test and Results and an external positioning system GPS. As a result, we studied the path of a drone using visual input that reduces GPS problems. And we found that a CNN with a fully connected regulator could successfully predict the steering angles needed to move the drone in a predetermined direction and steer the drone in the data path.
کلیدواژه ها:
Autopilot of a drone ، Automatic navigation a UAV ، intelligent navigation a UAV
نویسندگان
Reza Alipour
Researcher