Unmanned aerial vehicle power estimation with non-polarity charging system

• Main Author: Hadi, Al Obaidi Mohammed Rmaez
• Format: Thesis
• Language: English
• Published: 2019
• Subjects: Drone aircraft; Aerospace engineering
• Online Access: http://psasir.upm.edu.my/id/eprint/85340/1/FK%202020%204%20ir.pdf; http://psasir.upm.edu.my/id/eprint/85340/

This study focuses on the Unmanned Aerial Vehicles of lightweight electric vertical take-off and landing vehicles that also referred to as drones, are becoming increasingly popular in various important application areas such as surveillance, military applications, and monitor hazardous environments. One of the problems for using aerial robots is the relatively small duration of the flight because of limited power capability to achieve a long period of mission continually. The discharge capability which is the the way the battery is discharged and the charge storage capacity limitation of their lithium-ion battery can restrict their flight time endurance. Moreover, the autonomous landing and charging must be highly precise due to the lack of precision and accuracy of landing systems, Therefore, the utilization of an automatic drone’s charging station is highly desirable for those robots. This thesis addresses the aforementioned UAV power limitation problems and landing inaccuracy by Proposing a power management system that includes smart design of platform and charging circuit that can overcome the landing in an inaccurate manner disregarding to the UAV orientation, which is done by proposing a new platform topology. This charging system employed six bridge diodes in this topology to integrate with a platform configuration of a consecutive positive and negative charging plate. This integration of components has created a UAV charging system characterized by non-requirement for accurate landing to adjust the charging polarity when the UAV lands on the charging platform. However, Due to complications of flight maneuvers and highly unpredictable power consumption that can occur during flight missions, its necessary to establish a power monitoring system by measuring the real-time voltage and current of the UAV battery status and determine the flight time-efficient operation manner. However, as of today all the processes in the robotic system restrict in-air movement, consume energy and thereby defining the overall operation time limit, the problem is more pronounced in battery powered electric UAVs since different flight regimes like take off/landing and cruise have different power requirements and dead stick condition (battery depletion during flight mission) can have catastrophic consequences. the power rate measurements were conducted by developing a wireless monitoring circuit, which is proposed for this purpose that has the benefits of energy monitoring. The proposed wireless monitoring circuit characterized by its dependence on the RF-node embedded microcontroller to process and transmit the data to a monitoring station for more analysis. Therefore, this topology utilizes its own microcontroller and software driver to perform its functions. However, the unmonitored battery discharge is the major problem for UAVs to travel long distances, at least half of the energy in the battery must be saved to travel back to the launch site for recharging to address this issue, this work proposes an algorithm to analyze the energy data that collected through the proposed monitoring system to achieve modelling the power consumption patterns of particular drone trajectories with a specified environmental condition, which is solved by applying a robust approach called Bode Equation Vector Fitting (BEVF) algorithm and derive a measurement-based model for endurance estimation of such drone type. the results of the experimentation for a designing methodology that employed has successfully adjusted the polarity of the input power and improve the efficiency of the charging to 86%. Regarding the results of the wireless monitoring has successfully transferred the power data to the monitoring station. Moreover, BEVF algorithm has successfully reduced the RMS error equal to 1.93 in full discharge analysis between the measurement data and the modeling data in which is tuned to enhance the power modeling.