Student Projects
Contact Information
University: University of Leeds
Team Member(s): Joe Beaumont, Davis Chamia, Rhys Neild, Leo Rampen
Faculty Advisors: Rob Hewson, Martin Levesley, Peter Culmer
Email Address: leedsblimp@gmail.com
Country: UK
Project Information
Title:
Design of an Autonomous Airship Observation Platform
Description:
We are a team of 4 students from the University of Leeds, UK, currently involved in the design and manufacture of an airship that will incorporate autonomous navigation and buoyant technology to achieve high endurance flight for industrial inspection, surveillance and mapping applications.
Products:
sbRIO, LabView, IMAQdx
The Challenge:
The goal of this project is to demonstrate an autonomous airship for the purpose of low cost observation. Low cost observation platforms are becoming increasingly in demand. Whether it is for remote observation for civil law enforcement, mapping applications or powerline inspection. The project will deliver a demonstrator of the low cost observation platform for a payload of sensors (weighing in the region of 0.5kg) The system will have a degree of autonomy and ideally will have a degree of navigational capability (such as that provided through GPS and inertial sensors). The outcome of the project will be a flight demonstration of the system
The Solution:
Work has been proceeding to design and manufacture a blimp that will be capable of transmitting real time video data to the operator from an aerial location over long endurance flights. A helium envelope will hold 2.3 cubic meters of helium in order to give the aircraft buoyancy, accounting for 91% of the total weight. The propulsion system consists of 3 propellers in order to achieve a fast forward velocity as well as a high level of maneuverability. The control of this system will be made possible using a sbRIO housed in a gondola on the underside of the blimp. A high resolution IP camera will be used to capture video. This will analysed using NI IMAQdx so that the airship can navigate autonomously and compressed so that it can be transmitted and monitored by the user at the ground station. The sensor suite onboard the airship consists of an ultrasonic range finder to give accurate altitude near the ground, an inertial measurement unit to give orientation and a GPS module for position over a long range.
DESIGN
The design of the airship has come to fruition through the contributions of the individual team members in the following areas.
Joe has been responsible for the design and manufacture of the gas envelope and tail fins as well as the development of the flight physics section of the simulation. The envelope shape chosen has an efficient aerodynamic shape over a range of flight profiles whilst also having a low structural weight. Many unconventional designs were also considered but it was found they were not suitable for our specifications. The tail fins were sized such as to negate the unstable aerodynamic moment which would have been produced by a bare hull. The flight physics simulation consists of the 6 degree of freedom non-linear equations of motion for the airship with slender body theory used to develop expressions for the aerodynamic forces and moments.
Davis has been working on the design of the propulsion system to enable the blimp to maneuver effectively whilst minimizing the power consumption. This work involves testing a wide range of motor and propeller combinations and recording their performance in terms of thrust and efficiency. The propulsion layout consists of two main motors mounted on a single shaft to control surge and heave and a rear motor to provide pitch and yaw control.
Rhys' work has been divided between two distinct areas: the design of the gondola and the vision system. The gondola design calls for a structure that can house the components securely whilst minimizing drag and weight. The components are housed in two separate gondolas so that the centre of mass and center of thrust can be positioned independently. Industrial manufacturing techniques have been employed to build two gondolas that fulfil the structural, aerodynamic and mass requirements. The vision system aims to complement the navigation system to provide relative localization in unmapped environments. Using a lightweight IP camera, the opportunity is presented to fulfil the observation requirements, via a wireless communication link, and augment the autonomous navigation capabilities through image processing techniques.
Leo has designed the electronics and systems that enable communication between all the actuation and sensing hardware on board. This has required implementation of the I2C protocol for sensors, Pulse Width Modulation for motor and servo commands and a TCP IP protocol to enable communication with the ground station over WiFi. Leo has also been working on a control system so that active stabilization and positional control can be accomplished using several independent PID feedback loops.
SIMULATION
A simulation of the blimp has been established so that the control system can be tested.
The image above shows the blimp represented as an imported .stl graphic in the OpenGL API in Labview. Inputs are given by the user to control the orientation and position of the blimp. A real time controller generates motor inputs based on a PID feedback loop. These motor commands are used in a dynamic model along with aerodynamic effects to generate responses that define the vehicle state in each iteration of the loop. This simulation has aided the development of controller tuning by analyzing the responses such as the pitch orientation shown by the waveform graph. The scene is simulated from the camera perspective so that the software governing the camera gimbal can be tested. The window on the right shows image processing implemented to detect the position of the virtual sphere.
PROTOTYPING
This prototype was made to familiarize ourselves with manufacturing techniques with the polyurethane material. It is 1/4 scale of the length and 1/64 of the volume of the full scale blimp.
The electronics were also prototyped. This image shows the rangefinder, inertial measurement unit and GPS breakout board being interfaced with the sbRIO9602 for testing purposes.
MANUFACTURING
These images show the propulsion gondola assembly. The 1m length carbon rod can rotate 360 degrees to enable trust vectoring. Two electronic speed controllers are used to control the brushless motors on the ends of the rod. This propulsion system is complemented by a rear motor that is housed in the downward vertical fin to provide yaw and pitch control.
This image shows the main gondola used to hold the battery, the sbRIO and the sensor suite including an Axis IP camera mounted on a 2-axis gimbal. The base of the gondola is manufactured from a closed cell expanded polystyrene and the canopy is a vacuum formed High Impact polystyrene plastic. The weight of this assembly is approximately 1.2kg.
The helium envelope can hold 2300 liters of helium giving a buoyancy force of 2.3kg. The video below shows the envelopebeing inflated with air to check for leaks.
FUTURE WORK
We are currently working to integrate the dynamic modeling into our high-level software so that we can validate all the controls communication and navigation software in a simulated environment. Flight tests are to commence shortly which will enable us to improve the autonomy of the navigation system.
Thanks for reading. Please 'like'
