Quad-rotor Unmanned Aerial Vehicle
Quad-rotor Unmanned Aerial Vehicle
To create a self stabilising flying platform capable of autonomous navigation for aerial photography, tracking, and deployment of ground based security surveillance robots.
The UAV required a large thrust to weight ratio to enable it to carry small payloads of 250-500g. In addition to this carrying capacity very stable flight was necessary for capturing quality video and still imagery. The craft should have up to a kilometre of range. The remote control system should be able to receive smart commands such as GPS way points, photography and video instructions (orientation, altitude, panoramic shot sequencing etc), as well as land-deploy-takeoff etc. The interface for this needed to be secure but easily accessible for a wide range of scenarios.
The craft design had a number of physical constraints to be met, particularly the total weight of the UAV. By understanding interactions between the inertial properties of the craft and the control system a few mechanical design changes allowed the mathematical control system to be drastically simplified for much higher processing speed and greater dynamic response. The finalised design of the wing structure had to be very light weight but maintain a very high stiffness to suppress mechanical vibration and resist shock loading. This was achieved through the use of a complete CAD package including FEA (stress, strain, fatigue, vibration), motion simulation, CFD (propeller modelling), and CAM to produce the CNC aluminium structure. After developing the airframe the electrical system was broken down in the modules such as the motor control system, power management, intelligent battery management, sensory array, wireless link, and CPU. The subsystems were all integrated together on a common control bus (SPI) to allow flexibility for adding future features as well as minimising wiring and PCB space. All PCBs were designed from scratch to be plug and play on a common peripheral bus. Custom design allowed the boards to be condensed saving weight and space and gave the full flexibility needed. Finally the control system was implemented using fixed point filtering to dramatically increase speed and decrease code size. Only part of the system was required to be modelled in MATLAB for simulation and tuning with the remaining testing done online on the UAV hardware itself. A low cost 32 bit ARM core processor provided greater than 1kHz control system bandwidth and was tuned via a wireless link during run time.
The UAV proves that low cost sensors can be integrated using the latest embedded processors and signal processing techniques to meet very high performance requirements. The flexibility of firmware changes means that the systems can easily be re tasked to a wide range of markets creating endless possible applications.
• 32 bit ARM core embedded controller, fixed point signal processing at 80MHz
• Six axis inertial measurement unit (IMU) created by integrating a tri-axis accelerometer with three MEMS gyros via a special signal processing engine.
• Complete navigation system hardware consisting of a digital compass, ultrasonic ranger,
precision altimeter, motor BEMF sensing, custom differential GPS, and a variety of voltage
sensors to create a full sensory array.
• Wireless data logging or intermittent instruction updates
• Reliable operating range of 1.5km at 2.4GHz
<< Back to Case Studies