Application of neuro-fuzzy maximum solar power tracking controller to increase power profit of unmanned vehicles

This thesis presents a novel hybrid adaptive neuro-fuzzy control system that aims to extend the flight duration of multirotor vehicles by maximizing solar energy absorption throughout the day, from early morning until sunset. The system continuously monitors the maximum power point of the multirotor system using surface-mounted solar panels, enabling it to navigate with variable roll angles, ensuring optimal energy harvest without compromising flight stability or navigation performance. By dynamically adjusting its orientation in response to solar radiation, it can absorb more solar energy compared to traditional fixed solar systems, which typically collect limited power and do not add value to the multirotor vehicle's flight capabilities. This adaptive approach makes multirotor systems more efficient by providing uninterrupted power, allowing for longer flight endurance and data transfer, particularly in environments with fluctuating sunlight conditions. It also presents an advanced control scheme for the multirotor vehicle's integrated power system, specifically targeting the DC-DC converter responsible for voltage control and power transfer. The control architecture combines a PI controller optimized through particle swarm optimization with a neuro-fuzzy controller arranged in a cascade configuration. This design leverages the strengths of both control methods: The PI controller, with its minimum tuning parameters, provides fast and stable voltage regulation, while the neuro-fuzzy controller, which manages power and current fluctuations, adapts based on learned experience to improve overall efficiency. The particle swarm optimization process fine-tunes the PI parameters for optimal dynamic and steady-state responses, resulting in a system that can effectively adapt to changing environmental conditions, such as changing sunlight and weather patterns. Extensive testing has shown that this integrated control approach increases DC-DC converter efficiency by at least 15% compared to neuro-fuzzy control alone, while also significantly reducing tracking errors and improving power output stability. Overall, the combined use of adaptive solar tracking with advanced power management control strategies offers a robust solution for extending multi-rotor vehicle flight endurances, promoting sustainable energy use, and improving operational reliability in renewable energy-powered autonomous systems.