International Research Journal of Engineering and Technology (IRJET) Volume: 11 Issue: 05 | May 2024
www.irjet.net
e-ISSN: 2395-0056 p-ISSN: 2395-0072
Performance Prediction using Exploratory Data Analysis with the Attitude Parameters of a Quadcopter Darshit Dhake1, Himanshu Rawat2, Swapnil Mohanty3, Shardul Mehetar4 1-2Students, Dept. of Computer Science, Lokmanya Tilak College of Engineering, Navi Mumbai, Maharashtra, India 3Student, Department of Ceramic Engineering, National Institute of Technology Rourkela.
.4Student, Dept. of Computer Science, Terna Engineering College, Navi Mumbai, Maharashtra, India
---------------------------------------------------------------------***--------------------------------------------------------------------ATT parameter in the binary file. The binary file is fed as Abstract- Drones are widely used in the world for various input to the python code which contains the programming to extract data under sub parameters like des_roll, roll, des_pitch, pitch, des_yaw, yaw, time etc. We are more particular in understanding the patters of the actual and desired values of roll, pitch and yaw values. These attitude parameters are controlled using the quadcopter's motor speeds. By adjusting the speeds of the motors in various combinations, the quadcopter can achieve the desired roll, pitch, and yaw angles. To maintain stable flight, the attitude control system constantly monitors the quadcopter's orientation and makes adjustments as needed. The following process is followed under two different scenarios. One where there was a smooth flight during the mission and the second when there was a crash during the mission. The endurance of flight in both the cases is about 14.5 minutes. Exploratory Data Analysis is performed in both the cases and the performance of the flight is predicted further. This paper is divided into different sections. The drone was customized as described in [1]. The description of the paper gets initiated by considering the quadcopter dynamics as in [2]. This is described in section II. Section III talks about the software requirements for this project. Mission planner, python, UAV Log Viewer along with other packages need to be installed in the python environment in order to get the required results. Section IV, talks about the approach towards the project or the pseudo code of the execution of the python program. Section V talks about the discussions upon the results achieved. Further in section VI future works are discussed.
applications ranging from a drone racing and until surveillance purposes. The performance of a drone could be predicted using exploratory data analysis thereby they could be maintained for safe and reliable field operations. Various techniques are being conducted in the industry to predict the performance of the drone in the industries. One such analysis would be the using the attitude parameters collected from the SD card of any flight controller used in quadcopters. Exploratory Data Analysis is applied to the obtained parameters and results are further compared for prediction of safe operation and unsafe operations for the next flight. We perform post flight analysis, where in the information is obtained from data flash log files when loaded in the mission planner software. The extracted data is sent into EDA model which predicts the reliable operation by using the mode function in the dataframe.
Keywords-Drones, Exploratory Data Analysis, Performance prediction
I. INTRODUCTION The drones these days are integrated to several applications in the current scenario ranging from racing to surveillance. Incorporation of better performance is in high demand in these days when it comes to a fact of having full mission flights. In this paper, we talk about the attitude parameters that are out from a PD controller used in the quadcopter. The copter taken into the consideration is the micro drone customized based for the purpose of hobbyist flying. Here the intention of the pilot is to learn the drone piloting and hence it is important to know if the drone would fly the very next time. The drone has Ardupilot based Pixhawk cube flight controller and the parameters are accessed via the Mission Planner Software. Post flight analysis is done using the telemetry and data flash log files. Data flash logs are preferred as the communication between the air unit and ground unit remains established throughout the mission flight. The parameters from the dataflash log files cannot be viewed initially as they have .bin or binary format. In order to have a visualization of the parameters they are converted to .log format through the file convert options available in the mission planner software. Here as attitude analysis is performed, we are interested to look upon the data under © 2024, IRJET
|
Impact Factor value: 8.226
Quadcopters Flying machines or air vehicles are generally classified based on the way the wings are mounted. Hence an aerial vehicle could be either fixed wing, e.g., airplanes, or rotating wing, e.g., helicopters. A multi-rotor helicopter with the ability to VTOL is known as a quadrotor or quadcopter. Four rotors lift them and move them forward. They are classified as rotorcraft among aircraft since they are lifted and propelled by a group of rotors. Two identical fixed pitch propeller pairs, two rotating in the clockwise (CW) and two in the counter-clockwise (CCW) directions, are used by quad copters. These regulate lift and torque by varying RPM. Computer/electronic systems can regulate the motion of the vehicle by altering the torque load and |
ISO 9001:2008 Certified Journal
|
Page 2263