Abstract:
This project presents the design, development, and testing of a robotic system for cleaning solar panels. The primary objective was to create a robot that is highly efficient, lightweight, and cost-effective.
We opted for a complete aluminum extrusion frame due to its superior rigidity, lightweight design, and modularity. This choice optimized robot performance and weight efficiency. A 4-wheel tank drivetrain with individual motors was chosen for precise control, maneuverability, and traction. Additional idle wheels offer support and prevent falls. A helical microfiber cloth attached to a rotating shaft ensures efficient dust removal, adaptability to panel textures, and minimized wear and tear. An Arduino-based system controls robot movement, cleaning functions, and sensor data acquisition. Sensors: A Light Dependent Resistor (LDR) detects dust accumulation, enabling automated cleaning frequency based on machine learning. A regression classifier predicts cleaning needs using dust density, solar irradiance, and panel power output, reducing sensor requirements and optimizing operation. Comprehensive testing evaluated cleaning efficiency, speed, power consumption, and sensor accuracy. The robot achieved desired performance metrics.
This project successfully developed a highly efficient, lightweight, and cost-effective solar panel cleaning robot. The chosen design and functionalities contribute significantly to its success. Future work could focus on further sensor optimization and machine learning algorithm refinement.
