Canberra, Australia

Australian Defence Force Academy

www.adfa.edu.au
Canberra, Australia

The Australian Defence Force Academy is a tri-service military Academy that provides military and tertiary academic education for junior officers of the Australian Defence Force in the Royal Australian Navy , Australian Army and Royal Australian Air Force .Tertiary education is provided by the University of New South Wales Canberra campus, which is the awarding body for ADFA qualifications. Apart from educating future leaders of the Australian Defence Force, UNSW campus also provides postgraduate programs and short courses both to Department of Defence personnel and the general public.The stated purpose of ADFA is "to serve Australia by providing the Australian Defence Force with tertiary graduates who have the attributes, intellect and skills required of an officer."ADFA is located in the suburb of Campbell, Canberra, Australian Capital Territory, near the Australian Government district of Russell. It is situated next to Mount Pleasant, which gives some parts of ADFA a view over the rest of Canberra. The ADFA is also adjacent to the Australian Army military academy, the Royal Military College, Duntroon.The junior officers that attend the Australian Defence Force Academy hold the rank of Midshipman in the Royal Australian Navy, Officer Cadet in the Australian Army and Officer Cadet in the Royal Australian Air Force. The Academy is currently commanded by Air Commodore Alan Clements. Wikipedia.


Time filter

Source Type

Merrick K.E.,Australian Defence Force Academy
IEEE Transactions on Autonomous Mental Development | Year: 2012

Incorporating intrinsic motivation with reinforcement learning can permit agents to independently choose, which skills they will develop, or to change their focus of attention to learn different skills at different times. This implies an autonomous developmental process for skills in which a skill-acquisition goal is first identified, then a skill is learned to solve the goal. The learned skill may then be stored, reused, temporarily ignored or even permanently erased. This paper formalizes the developmental process for skills by proposing a goal-lifecycle using the option framework for motivated reinforcement learning agents. The paper shows how the goal-lifecycle can be used as a basis for designing motivational state-spaces that permit agents to reason introspectively and autonomously about when to learn skills to solve goals, when to activate skills, when to suspend activation of skills or when to delete skills. An algorithm is presented that simultaneously learns: 1) an introspective policy mapping motivational states to decisions that change the agent's motivational state, and 2) multiple option policies mapping sensed states and actions to achieve various domain-specific goals. Two variations of agents using this model are compared to motivated reinforcement learning agents without introspection for controlling non-player characters in a computer game scenario. Results show that agents using introspection can focus their attention on learning more complex skills than agents without introspection. In addition, they can learn these skills more effectively. © 2009-2011 IEEE.


Maalouf A.I.,Australian Defence Force Academy | Petersen I.R.,Australian Defence Force Academy
IEEE Transactions on Automatic Control | Year: 2011

This paper considers the bounded real properties for a class of linear quantum systems which can be defined by complex quantum stochastic differential equations in terms of annihilation operators only. The paper considers complex quantum versions of the Bounded Real Lemma, the Strict Bounded Real Lemma and the Lossless Bounded Real Lemma. For the class of quantum systems under consideration, it is shown that the question of physical realizability is related to the Bounded Real and Lossless Bounded Real properties. © 2006 IEEE.


Maalouf A.I.,Australian Defence Force Academy | Petersen I.R.,Australian Defence Force Academy
IEEE Transactions on Automatic Control | Year: 2011

This paper considers a coherent H∞ control problem for a class of linear quantum systems which can be defined by complex quantum stochastic differential equations in terms of annihilation operators only. For this class of quantum systems, a solution to the H∞ control problem can be obtained in terms of a pair of complex algebraic Riccati equations. In addition, the question of physical realizability of the resulting quantum controllers is related to a bounded real property. © 2006 IEEE.


This paper presents a new approach to constructive output feedback robust nonlinear guaranteed cost controller design. The approach involves a class of controllers which include copies of the slope bounded nonlinearities occurring in the plant. Dynamic multipliers are introduced to exploit these repeated nonlinearities. The linear part of the controller is synthesized using minimax LQG control theory. © 2010 Elsevier Ltd. All rights reserved.


Petersen I.R.,Australian Defence Force Academy
IEEE Transactions on Automatic Control | Year: 2013

This technical note considers the use of singular perturbation approximations for a class of linear quantum systems arising in the area of linear quantum optics. The technical note presents results on the physical realizability properties of the approximate system arising from singular perturbation model reduction. © 2012 IEEE.


Petersen I.R.,Australian Defence Force Academy
Systems and Control Letters | Year: 2012

This paper presents a method for approximating a class of complex transfer function matrices corresponding to physically realizable complex linear quantum systems. The class of linear quantum systems under consideration includes interconnections of passive optical components such as cavities, beam-splitters, phase-shifters and interferometers. This approximation method builds on a previous result for cascade realization and gives good approximations at low frequencies. © 2011 Elsevier B.V. All rights reserved.


Pickering M.R.,Australian Defence Force Academy
Proceedings - International Conference on Image Processing, ICIP | Year: 2011

Multi-modal similarity measures are required to register images of the same object using different sensors. This registration is often required for medical images of the same patient captured using different imaging modalities such as MRI, CT and PET. In this paper, a new multi-modal similarity measure is proposed which is based on calculating the sum-of-conditional variances from the joint histogram of the two images to be registered. The formulation of this new similarity measure allows the standard Gauss-Newton optimization procedure to be used. Our experimental results show that this new approach is more accurate and robust than the most common and best performing alternative and is also more computationally efficient. © 2011 IEEE.


Ugrinovskii V.,Australian Defence Force Academy
Automatica | Year: 2013

The paper considers a distributed robust estimation problem over a network with Markovian randomly varying topology. The objective is to deal with network variations locally, by switching observer gains at affected nodes only. We propose sufficient conditions which guarantee a suboptimal H∞ level of relative disagreement of estimates in such observer networks. When the status of the network is known globally, these sufficient conditions enable the network gains to be computed by solving certain LMIs. When the nodes are to rely on a locally available information about the network topology, additional rank constraints are used to condition the gains, given this information. The results are complemented by necessary conditions which relate properties of the interconnection graph Laplacian to the mean-square detectability of the plant through measurement and interconnection channels. © 2012 Elsevier Ltd. All rights reserved.


Shaiju A.J.,Indian Institute of Technology Madras | Petersen I.R.,Australian Defence Force Academy
IEEE Transactions on Automatic Control | Year: 2012

A recently emerging approach to the feedback control of linear quantum systems involves the use of a controller which itself is a quantum linear system. This approach to quantum feedback control, referred to as coherent quantum feedback control, has the advantage that it does not destroy quantum information, is fast, and has the potential for efficient implementation. An important issue which arises both in the synthesis of linear coherent quantum controllers and in the modeling of linear quantum systems, is the issue of physical realizability. This issue relates to the property of whether a given set of linear quantum stochastic differential equations corresponds to a physical quantum system satisfying the laws of quantum mechanics. Under suitable assumptions, the paper shows that the question of physical realizability is equivalent to a frequency domain (J,J) -unitary condition. This is important in controller synthesis since it is the transfer function matrix of the controller which determines the closed loop system behavior. © 2012 IEEE.


Ugrinovskii V.,Australian Defence Force Academy
Proceedings of the IEEE Conference on Decision and Control | Year: 2011

The paper considers a problem of consensus-based synchronization of uncertain parameter varying multi-agent systems with Lipschitz-continuous nonlinearities. The objective is to construct simultaneously consensus and observer schedules for each agent to ensure an asymptotic synchronized behaviour of all agents. A gain-scheduling algorithm is proposed which solves this problem while maintaining a specified suboptimal H ∞ level of relative disagreement between the agents. The algorithm uses interpolation to ensure the continuity of the interconnection and observer gains. It preserves the H ∞ consensus properties of the interpolants. © 2011 IEEE.

Loading Australian Defence Force Academy collaborators
Loading Australian Defence Force Academy collaborators