One of many challenges in swarm robot coordination is
to generate geometric patterns from the robots using a decentralized control approach. Such formations have numerous applications ranging from military to medical and space to underwater. This approach uses artificial force based controller which navigates the robots in a decentralized manner. The mathematical
analysis of the controller for stability and cohesiveness provides a criterion for selecting the weighing parameters for the controller. Moreover, the extendability of the concept of artificial force based control of a swarm is demonstrated in an application specific scenario. Here, a two-stage controller which satisfy special control requirements of an airborne guided weapon system was also derived.
MATLAB simulation results
The simulation case studies developed in MATLAB to evaluate the theoretical assertions are provided as video clips. Click on the play button of each clip to activate and play the video.
Demonstrating the formation of multi-agent system into a pre-defined shape
Addition of new members to an already stabilized formation. (Scalability of the system)
Obstacle avoiding behviour of the robotic swarm. Initially the robots are in a line formation, and then converges to the pre-defined shape contour while avoiding two obstacles defined as contours.
Publications
Samitha W. Ekanayake, “Formation of
networked mobile robots”, 2009, PhD Thesis, School of Engineering, Deakin University, Australia. [pdf]
Samitha W. Ekanayake and Pubudu N. Pathirana, “Formations of Robotic Swarm - An Artificial Force Based Approach,” International Journal of Advanced Robotic Systems, Vol. 6, No. 1 (2009) pp. 7-24 [link]
Ekanayake S.W. and Pathirana P.N., “Two Stage Architecture for Navigating Multiple
Guided Weapons into a Widespread Target,” 2008 IEEE Aerospace Conference, pp.1-
20, March 2008, USA.
Ekanayake S.W. and Pathirana P.N., “Geometric formations in swarm aggregation: An
artificial formation force based approach,” Third International Conference on Information
and Automation for Sustainability, 2007. ICIAFS 2007, pp.82-87, 4-6 Dec. 2007
Ekanayake S.W. and Pathirana P.N., “Smart Cluster Bombs - Control of Multi-agent
Systems for Military Applications,” Networking, Sensing and Control, 2007 IEEE
International Conference on , pp.471-476, 15-17 April 2007
Ekanayake S.W. and Pathirana P.N., “Artificial Formation Forces for Stable Aggregation
of Multi-Agent System,” Information and Automation, 2006. ICIA 2006. International
Conference on , pp.129-134, 15-17 Dec. 2006
Optimization based approach
Summary
Formation of autonomous mobile robots to an arbitrary geometric pattern in a distributed fashion is a fundamental problem in formation control. This paper presents a new fully distributed, memoryless (oblivious) algorithm to the underlying problem via distributed optimization techniques. The optimization minimizes an appropriately defined difference function between the current robot distribution and target geometric pattern. The optimization processes are performed independently by individual robots in their local coordinate system. A movement strategy derived from the results of the distributed optimizations guarantees that every movement makes the current robot configuration approaches the target geometric pattern until the final pattern is reached.
Publications
Zhang, H., Pathirana,P.N., “Optimization Based Formation of Autonomous Mobile Robots”, Robotica, (Accepted)
Vision-based tracking of an object using perspective projection inherently results in non-linear measurement equations in the Cartesian coordinates. The underlying object kinematics can be modelled by a linear system. In this research we introduced a measurement conversion technique that analytically transforms the non-linear measurement equations obtained from a stereo-vision system into a system of linear measurement equations. We then design a robust linear filter around the converted measurement system. The state estimation error of the proposed filter is bounded and we provide a rigorous theoretical analysis of this result. The performance of the robust filter developed in this paper is demonstrated via computer simulation and via practical experimentation using a robotic manipulator as a target. The proposed filter is shown to outperform the extended Kalman filter (EKF).
Publications
Pubudu N. Pathirana, Adrian N. Bishop, Andrey V. Savkin, Samitha W. Ekanayake and Timothy J. Black, “A method for stereo-vision-based tracking for robotic applications,”Robotica, 2010, Vol. 28, No. 4, Cambridge University Press, pp 517-524 (doi:10.1017/S0263574709005827) [link]
Pathirana, Pubudu, Lim, A., Savkin, Andrey and Hodgson, Peter (2007) Robust video/ultrasonic fusion-based estimation for automotive applications, IEEE transactions on vehicular technology, vol. 56, no. 4, pp. 1631-1639, Institute of Electrical and Electronics Engineers (IEEE), Piscataway, N.J.
Pathirana Pubudu N., Bishop Adrian N., Savkin Andrey V., Ekanayake Samitha W. and
Black Timothy J., “A method for stereo-vision based tracking for robotic applications,”
Decision and Control, 2008. CDC 2008. 47th IEEE Conference on , pp.1298-1303, 9-11
Dec. 2008
Pathirana, Pubudu, Lim, Allan, Carminati, John and Premaratne, Malin (2007) Simultaneous estimation of optical flow and object state: A modified approach to optical flow calculation, 2007 IEEE International Conference on Networking, Sensing, and Control, pp. 634-638, Institute of Electrical and Electronics Engineers (IEEE), Piscataway, N.J
Stereo-vision based SLAM in robotic applications
Summary
The problem of visual sumultaneous localization and mapping (SLAM) is investigated in this research using the ideas and algorithms from robust control and estimation theory. Using a stereo-vision based sensor, a nonlinear measurement model is derived which leads to nonlinear measurements of the landmark coordinates along with optical flow based measurements of the relative robot-landmark velocity. Using a novel analytical measurement trasnformation, the nonlinear SLAM problem is converted into the linear domain and solved using a robust linear filter.
Publications
Pathirana, P.N., Savkin,A.V., Ekanayake, S.W., Bauer, N.J., “A Robust Solution to the Stereo-Vision Based Localization and Mapping Problem with Steady and Moving Landmarks”, Advanced Robotics, Scheduled to appear in Vol. 25, No. 5 (April 2011)
Pathirana,P.N.,Bishop, A.N.,Savkin, A.V.,Ekanayake,S.W.,Bauer, N., “A robust set-valued state estimation approach to the problem of vision based SLAM for mobile robots”, The 10th EUCA Series of European Control Conferences (ECC),Budapest, Hungary, 23-26 August 2009
Experimental setup: Two cameras mounted on a mobile robot is used to perform the exeriments in this research project
Bishop, Adrian, Pathirana, Pubudu and Savkin, Andrey* (2007) Radar target tracking via robust linear filtering, IEEE signal processing letters, vol. 14, no. 12, pp. 1028-1031, Institute of Electrical and Electronics Engineers (IEEE), Piscataway, N.J.
Bishop, Adrian and Pathirana, Pubudu (2007) Localization of emitters via the intersection of bearing lines: a ghost elimination approach, IEEE transactions on vehicular technology, vol. 56, no. 5, pp. 3106-3110, IEEE, Piscataway, N.J.
Localizing a tranmitter based on various measurements in a distributed sensor network is an important aspect in modern ad-hoc and mobile wireless communication platforms. We have investigated several approaches in the tranmitter localization problem: Received signal strength and Time Difference of Arrival (TDoA) as well as developed sensor placement strategies to perform optimal localization of an emitter.
Publications
Rolfe B.F., Ekanayake S.W., Pathirana P.N. and Palaniswami M., “Localization with
orientation using RSSI measurements: RF map based approach,” Intelligent Sensors,
Sensor Networks and Information, 2007. ISSNIP 2007. 3rd International Conference on ,
pp.311-316, 3-6 Dec. 2007
Beladi, Somaieh and Pathirana, Pubudu N. (2008) TDOA based transmitter localization with minimum number of teceivers and power measurements, ICARCV 2008 : 10th International Conference on Control, Automation, Robotics & Vision, pp. 1259-1264, IEEE, Piscataway, N.J.
Bishop, A.N.*, Fidan, B., Anderson, B.D.O., Dogancay, K. and Pathirana, P. (2008) Optimal range-difference-based localization considering geometrical constraints, IEEE Journal of Oceanic Engineering, vol. 33, no. 3, pp. 289-301, IEEE Journal of Oceanic Engineering, N.Y., N.Y.
Beladi, Somaieh, Pathirana, Pubudu and Hodgson, Peter (2007) Planar receiver placement for unique emitter localization for indoor applications, 3rd IEEE International Conference on Wireless and Mobile Computing, Networking and Communications, pp. 1-6, Institute of Electrical and Electronics Engineers (IEEE), Piscataway, N.J.
Transmission power control for energy conservation and Quality of Service maintenance in wireless networks.
Many wireless ad-hoc networks, such as sensor networks and military communication networks, are inherently associated with restrictions in power consumption, mainly due to the limited energy resources such as batteries. In our research group we have investigated transmitter power control strategies to maximize the network time of the sensor devices as well as to enable effective and high-quality communication between the communication devices. Not limiting the research into the sensor networking domain, we have extended the research into the tranmission power control in CDMA cellular networks.
Publications
Samitha W. Ekanayake, Pubudu N. Pathirana, “RSS Based Technologies in Wireless Sensor Networks”, Mobile and Wireless Communications - Network Layer and Circuit Level Design, INTECH, 2010, ISBN 978-953-307-042-1, pp 37-59 [link]
Ekanayake S.W., Pathirana P.N., Rolfe B.F. and Palaniswami M., “Energy Efficient,
Fully-Connected Mesh Networks for High Speed Applications,” IEEE Vehicular
Technology Conference, VTC Spring 2008., pp.2606-2610, 11-14 May 2008
Ekanayake S.W., Pathirana P.N. and Palaniswami M., “Maintaining Optimal Co-Channel
Interference for Power Efficient Wireless Communication,” Intelligent Sensors, Sensor
Networks and Information, 2007. ISSNIP 2007. 3rd International Conference on , pp.25-
30, 3-6 Dec. 2007
Zhang Huan, Pathirana Pubudu and Ekanayake Samitha, “Power control in CDMA
systems based on channel identification,” Intelligent Sensors, Sensor Networks and
Information Processing, 2008. ISSNIP 2008. International Conference on , pp.399-404,
15-18 Dec. 2008
Huan Zhang, Pathirana P. and Ekanayake S., “Robust Power Controllers in Cellular
Radio Systems,” Vehicular Technology Conference, 2008. VTC Spring 2008. IEEE ,
pp.2162-2166, 11-14 May 2008
BigNET generic sensor network testbed
The major contribution of the generic sensor network platform is the plug-and-play capability and the modular architecture, which enable a sensor node to be equipped with any type of sensor/actuator combination and wireless communication device (see above figure). It has been designed to accommodate up to 16 different sensor/actuator combinations at a time, and two communication and processing modules. This modular architecture allows the users to configure the sensor node in various configurations. The front-end of the sensor network is the internet based data visualization and device control/configuration tool. The internet based system allows the users to access the system around the globe via any web enabled device (such as PC, net book, PDA or a smart phone). Currently we have the system in three different configurations: position tracking system, environmental monitoring system and a web based remote device controller. In the position tracking system, the sensor node is equipped with a GPS sensor and a GSM communication module where as in the environmental monitoring system; we have integrated several other sensors (pH, temperature, water level) in addition to the GPS sensor. In the remote device control terminal, as opposed to the sensor modules used in previous two scenarios, we have equipped the sensor node with an actuator module that can control an electric load.
Click here to visit the BigNET testbed website hosted by Deakin University.
This project was an integrated sensing and networking platform development in collaboration with Melbourne University.
Publicatons:
Aravinda S Rao, Davood Lazadi, Reuben F. Tellis, Samitha W. Ekanayake and Pubudu
N. Pathirana, “Data Monitoring Sensor Network for BigNet Research Testbed”, Intelligent
Sensors, Sensor Networks and Information Processing, 2009. ISSNIP 2009.
International Conference on, Dec. 2009
Diffuison weighted Magnetic resonance imaging (DW MRI) is widely used to investigate the orientation and connectivity of the WM fibres. The majority of applications of DW MRI is to diagnosing and assessing various neurological disorders (e.g., Acute ischemia, Alzheimer disease) as well as behaviour and cognitive disorders (e.g., Schizophrenia, ADD, ADHD). The other important application of DW MRI is the brain tissue segmentation which has a key role in studying the structure and function of the brain. The most important segmentation method is to reconstruct the fibre tracts in the white matters of the brain non-invasively, which is called Tractography. For this purpose, the first step is to estimate the fibre orientation in each voxel in the brain. Different methods have been proposed to solve this problem such as diffusion tensor imaging (DTI), q-space imaging (QSI) and high angular resolution diffusion imaging (HARDI). Among the HARDI methods, Qball imaging was presented as a model free, linear and sensitive to multimodal diffusion approach to reconstruct diffusion orientation distribution function using HARDI signal. In our current study we concentrate on the Qball imaging to modify it by improving the required estimation methods and imposing the required physical constraint.
Functional magnetic resonance imaging (fMRI) is a non-invasive method to assess cortical activation of the human brain for a given stimulation by measuring changes in oxidation and regional blood flow. As movies are to still photography: functional magnetic resonance imaging (fMRI) is to structural MRI. Among all the fMRI methods, Blood Oxygenation Level Dependent (BOLD) contrast is the dominant technique. The BOLD signal is sensitive to the local deoxyhemoglobin concentration; however the BOLD signal does not directly measure the neural activity itself. Instead the BOLD effect is sensitive to a series of physiological responses that are referred collectively as the hemodynamic response for activation. Analysis of the fMRI data is more generally carried out in a General Linear Model (GLM) framework. But with the recent work showing the nonlinearity of the BOLD
response, several researchers have attempted to handle nonlinear characterization for these underlying brain processes. As a result parameterised models have also been developed for to model the nonlinear signal transduction. In the present study we use the most recent proposed such model and formulate it in the standard system matrix formulation. Then a parameter estimation and state estimation is proposed on the model by considering the actual colored and correlated noise structure in fMRI data, apart from the traditional white noise assumption. A maximum likelihood estimation is proposed on the parameter estimation of the noisy data. With previous applications of the Robust kalman filtering showing better performance on similar models compared to the other nonlinear filtering techniques such as EKF and PF, a Robust kalman filtering approach is proposed for the state estimation. Subsequent to the estimations, formulation of a generalised measure for the diagnose of Parkinsons desease is proposed with the nonlinear model.
