Friday, 17 September 2010

Publications update

I've updated my publications page to include our most recent conference and journal papers.

Saturday, 24 July 2010

Farnborough Air Show

Just back from exhibiting at Futures Day at the Farnborough Air Show.  Among the people we met, it was encouraging to bump into so many Bristol graduates in the aerospace business.  Now, watch this space as we work on our autonomous mini-Blades display team...

Wednesday, 21 July 2010

Congratulations Gillian

Gillian Clare (nee Keith) has passed her PhD viva with only minor corrections to be done.  Gillian's work on aircraft taxi optimization, supported by Airbus, identified a possible saving of total delay time of 56% at London's Heathrow Airport.  Gillian now moves on to work at BAE Systems in Filton.

Monday, 12 July 2010


Congratulations to all our graduates this year.

Special mention to Alison Eele, who completed her PhD with me earlier this year and graduated with the other PhDs on Friday.  Alison's work developed a novel path optimization technique combining global search with nonlinear dynamics constraints for obstacle avoidance.  Alison has now moved on to a postdoctoral research post at Cambridge working on air traffic management.

Monday, 5 July 2010

IMAV at RAF Waddington Airshow

I attended the RAF Waddington Airshow to exhibit research work on the DSTL stand in the UAV hangar.  In particular, the Intelligent Micro Air Vehicle (IMAV) was on static display (right).  (The cable and hanging wire were just for the exhibition - normally it is wireless and free-flying.)

The IMAV project combines Visual Simultaneous Localisation and Mapping (VSLAM) with guidance and control techniques to provide an autonomous indoor surveillance capability.  Users such as fire and rescue responders or police would use it for first investigations in hazardous situations.  It navigates solely by its own vision, making it independent of GPS, and does not require a skilled pilot.  Ince the research is focussed on the algorithms for control and navigation, the development hardware is deliberately simple: we use a modified radio-controlled toy helicopter and a commercial Wi-Fi webcam, again modified to suit SLAM.

As well as discussing UAV technology and research with a diverse audience, the airshow gave an opportunity to get up close with some UAV hardware.  For example, here's a picture of my car next to Watchkeeper (left).  Not much street cred I admit, but it's always difficult to get a sense of scale from UAV photos.

Thanks to Chris Jones from DSTL for giving us the opportunity to exhibit.

Background: IMAV is a joint project between the Departments of Computer Science and Aerospace Engineering at the University of Bristol.  It is funded by PhD studentships from the Faculty of Engineering with initial support from DSTL.  PhD students John Bartholemew (CS) and Kieran Wood (AE) are supervised by Drs. Andrew Calway (CS), Walterio Mayol-Cuevas (CS), Tom Richardson (AE) and Arthur Richards (AE).

Monday, 28 June 2010

Journal Paper: Distributed model predictive control of linear systems with persistent disturbances

Paul Trodden and Arthur Richards, Distributed model predictive control of linear systems with persistent disturbances, International Journal of Control, Online, published 24th June 2010

Abstract This article presents a new form of robust distributed model predictive control (MPC) for multiple dynamically decoupled subsystems, in which distributed control agents exchange plans to achieve satisfaction of coupling constraints. The new method offers greater flexibility in communications than existing robust methods, and relaxes restrictions on the order in which distributed computations are performed. The local controllers use the concept of tube MPC - in which an optimisation designs a tube for the system to follow rather than a trajectory - to achieve robust feasibility and stability despite the presence of persistent, bounded disturbances. A methodical exploration of the trades between performance and communication is provided by numerical simulations of an example scenario. It is shown that at low levels of inter-agent communication, distributed MPC can obtain a lower closed-loop cost than that obtained by a centralised implementation. A further example shows that the flexibility in communications means the new algorithm has a relatively low susceptibility to the adverse effects of delays in computation and communication.

Keywords: linear systems; distributed control; constrained control

DOI: 10.1080/00207179.2010.485280

Monday, 21 June 2010

Flight Control... Automated

We tried out one of our routing algorithms on a version of the game Flight Control.  When each new aircraft enters, a Mixed Integer Linear Program (MILP) optimization finds its best route to its target, avoiding the routes already set for other aircraft.  So far its best score is 165.  It tends to fail when aircraft don't quite follow the route they've been set, or if two get close to the runway entrance at the same time.  We're working on improving performance, but there's a catch: the more detail you put in the optimizer, the slower it responds.  That puts aircraft further off their assigned paths and leads to problems later on. Thanks to Colin for coding the game.

Monday, 1 February 2010

Congratulations to Paul on his PhD

Paul Trodden passed his PhD viva in August and has now completed the few corrections asked. Paul's PhD, sponsored by BAE Systems and an EPSRC CASE award, looked at robust distributed model predictive control (MPC). In particular, he developed ways to promote cooperation between agents, instead of greediness.

Paul will be staying on to work on a new ESA-funded project on MPC for spacecraft rendezvous.

Aerospace PhD graduates in February 2010
Left to right: Cormac McFarlane, Chris Jones, and their supervisor Tom Richardson; Arthur Richards, Djamel Rezgui, Paul Trodden and Jon DuBois

More posts about distributed MPC