Approximation of non-autonomous dynamic systems by continuous time recurrent neural networks

This work provides a framework for the approximation of a dynamic system of the form x˙=f(x)+g(x)u by dynamic recurrent neural network. This extends previous work in which approximate realisation of autonomous dynamic systems was proven. Given certain conditions, the first p output neural units of a dynamic n-dimensional neural model approximate at a desired proximity a p-dimensional dynamic system with n>p. The neural architecture studied is then successfully implemented in a nonlinear multivariable system identification case study.

Output envelope adjustment for an environmental fuzzy logic controller

A two-level fuzzy logic controller for use in air-conditioning systems is outlined in this paper. At the first level a simplified controller is produced from expert knowledge and envelope adjustment is introduced, while the second level provides a means for adapting this controller to different working spaces. The mechanism for adaption is easily implemented and can be used in real time. A series of simulations is presented to illustrate the proposed schema.

Quasi-decoupled fuzzy logic controller

The authors describe the design of a fuzzy logic controller for the control of a planar two-link manipulator. The plant is quasi-decoupled with respect to gravity. Complete decoupling is not achieved due to the nonoptimal nature of the expert rules. The performance of the fuzzy controller is compared to that of the critically damped computed torque controller. Results are presented complete with robustness tests.

Application of a hybrid tracking algorithm to motion analysis

An algorithm for tracking multiple feature positions in a dynamic image sequence is presented. This is achieved using a combination of two trajectory-based methods, with the resulting hybrid algorithm exhibiting the advantages of both. An optimizing exchange algorithm is described which enables short feature paths to be tracked without prior knowledge of the motion being studied. The resulting partial trajectories are then used to initialize a fast predictor algorithm which is capable of rapidly tracking multiple feature paths. As this predictor algorithm becomes tuned to the feature positions being tracked, it is shown how the location of occluded or poorly detected features can be predicted. The results of applying this tracking algorithm to data obtained from real-world scenes are then presented.

Distributed fuzzy logic for building management systems using local operating networks

This paper describes the development of an experimental distributed fuzzy control system for heating and ventilation (HVAC) systems within a building. Each local control loop is affected by a number of local variables, as well as information from neighboring controllers. By including this additional information it is hoped that a more equal allocation of resources can be achieved.

Learning in neural networks and stochastic approximation methods with averaging

The problem of adjusting the weights (learning) in multilayer feedforward neural networks (NN) is known to be of a high importance when utilizing NN techniques in various practical applications. The learning procedure is to be performed as fast as possible and in a simple computational fashion, the two requirements which are usually not satisfied practically by the methods developed so far. Moreover, the presence of random inaccuracies are usually not taken into account. In view of these three issues, an alternative stochastic approximation approach discussed in the paper, seems to be very promising.

Missing the logic of the text: Lord Butler’s report on intelligence on Iraqi weapons of mass destruction

This article focuses on the final report of Lord Butler’s review of British intelligence on weapons of mass destruction (WMD), specifically on its treatment of the accuracy of the use of intelligence on Iraqi WMD in a government dossier published in September 2002 ahead of the 2003 Iraq war. In the report, the demonstration of the accuracy of the “September Dossier” hinges on the insertion of tables that compare side-by-side quotations from this document and from intelligence assessments. The analysis of the textual and visual methods by which the report is written reveals how the logic of the comparative tables is missed in the Butler report: the logic of these tables requires that the comparison between quotations from the two documents should be performed at the level of their details but the Butler report performs its comparison only at a broad and general level.

Redefining the outer boundaries of EU Law: the Zambrano, McCarthy and Dereci trilogy

In its three recent rulings in the cases of Zambrano, McCarthy, and Dereci, the Court appears to have been determined to redefine the external boundaries of EU law, in cases involving the family reunification rights of Union citizens.These three judgments can be read as an indication that for Article 20 TFEU to apply, there is no longer a requirement of a cross-border element on the facts of the case, and that it is sufficient if the contested national measure has the effect of ‘depriving citizens of the Union of the genuine enjoyment of the substance’ of their rights (the ‘Zambrano principle’).The cases can, at the same time, also be read as a confirmation that the free movement provisions do – still – require a cross-border element and, in particular, the exercise of inter-State movement, in order to apply. Though the result in these cases has not been entirely unexpected, especially in the aftermath of the Rottmann ruling, it is rather problematic in that, although it is obvious that the Court wishes to redraw the line dividing the national and EU spheres of competence, it does not make it entirely clear where this line now lies and leaves many essential questions unanswered, which will obviously require some time to be resolved. EU lawyers are consequently, once more, left with having to decipher as best as they can the real intentions of the Court in this new line of case-law, which has been further complicated by the fact that what the Court seems to have given with one hand in Zambrano (and before that in Rottmann), has taken it back to a large extent through its rulings in McCarthy and Dereci, which appear to confine the former two cases to their own exceptional facts.6 Moreover, the ‘reverse discrimination Pandora’s box’, the opening of which appears to have been the real target of these references, remains untouched: instead of providing a direct solution to this problem, the Court has chosen to – once again – broaden the scope of the Treaty provisions in order to include within it as many situations as possible and, thus, prevent the emergence of this type of differential treatment on a case-by-case basis.As will be explained, nonetheless, this is by no means an appropriate solution to the reverse discrimination conundrum.

The GRAPE aerosol retrieval algorithm

The aerosol component of the Oxford-Rutherford Aerosol and Cloud (ORAC) combined cloud and aerosol retrieval scheme is described and the theoretical performance of the algorithm is analysed. ORAC is an optimal estimation retrieval scheme for deriving cloud and aerosol properties from measurements made by imaging satellite radiometers and, when applied to cloud free radiances, provides estimates of aerosol optical depth at a wavelength of 550 nm, aerosol effective radius and surface reflectance at 550 nm. The aerosol retrieval component of ORAC has several incarnations – this paper addresses the version which operates in conjunction with the cloud retrieval component of ORAC (described by Watts et al., 1998), as applied in producing the Global Retrieval of ATSR Cloud Parameters and Evaluation (GRAPE) data-set.