Input/output linearization using dynamic recurrent neural networks

This paper brings together two areas of research that have received considerable attention during the last years, namely feedback linearization and neural networks. A proposition that guarantees the Input/Output (I/O) linearization of nonlinear control affine systems with Dynamic Recurrent Neural Networks (DRNNs) is formulated and proved. The proposition and the linearization procedure are illustrated with the simulation of a single link manipulator.

Dynamic input/output linearization using recurrent neural networks

A dynamic recurrent neural network (DRNN) is used to input/output linearize a control affine system in the globally linearizing control (GLC) structure. The network is trained as a part of a closed loop that involves a PI controller, the goal is to use the network, as a dynamic feedback, to cancel the nonlinear terms of the plant. The stability of the configuration is guarantee if the network and the plant are asymptotically stable and the linearizing input is bounded.

Stable linearization using multilayer neural networks

The main limitation of linearization theory that prevents its application in practical problems is the need for an exact knowledge of the plant. This requirement is eliminated and it is shown that a multilayer network can synthesise the state feedback coefficients that linearize a nonlinear control affine plant. The stability of the linearizing closed loop can be guaranteed if the autonomous plant is asymptotically stable and the state feedback is bounded.

Reinterpreting ‘generic structure’: evolution, application and limitations of a concept

This paper traces the evolution of thegeneric structure concept in system dynamics and discusses the different practical uses to which they have been put. A review of previous work leads to the identification of three different views of what a ‘generic structure’ is and, hence, what transferability means. These different views are distinguishable in application as well as in theory. Examination of these interpretations shows that the assumptions behind them are quite distinct. From this analysis it is argued that it is no longer useful to treat ‘generic structure’ as a single concept since the unity it implies is only superficial. The conclusion is that the concept needs unbundling so that different assumptions about transferability of structure can be made explicit, and the role of generic structures as generalisable theories of dynamic behaviour in system dynamics theory and practice can be debated and clarified more effectively.

Reinterpreting 'generic structure': evolution, application and limitations of a concept

This paper traces the evolution of the generic structure concept in system dynamics and discusses the different practical uses to which they have been put. A review of previous work leads to the identification of three different views of what a generic structure is and, hence, what transferability means. These different views are distinguishable in application as well as in theory. Examination of these interpretations shows that the assumptions behind them are quite distinct. From this analysis it is argued that it is no longer useful to treat generic structure as a single concept since the unity it implies is only superficial. The conclusion is that the concept needs unbundling so that different assumptions about transferability of structure can be made explicit, and the role of generic structures as generalisable theories of dynamic behaviour in system dynamics theory and practice can be debated and clarified more effectively.

Dayside moving auroral transients related to LLBL dynamics

The NOAA-12 satellite skimmed through a region of dayside auroral activity over Svalbard on January 12, 1992. A sequence of auroral forms from two separated onset sites in the postnoon sector drifted westward towards magnetic noon. The auroral forms were associated with a population of injected magnetosheath plasma mixed with a secondary component of magnetospheric ions (>30 keV) that is a key signature of the low-latitude boundary layer (LLBL). The direction of motion of the cleft auroral forms and the basic features of the NOAA particle spectrograms indicate that the transients are related to LLBL on open field lines. The auroral transients are consistent with footprints of reconnection at the dayside magnetopause which is both patchy in space and sporadic in time.

Potential role of vegetation feedback in the climate sensitivity of high-latitude regions: a case study at 6000 years B.P.

Previous climate model simulations have shown that the configuration of the Earth's orbit during the early to mid-Holocene (approximately 10–5 kyr) can account for the generally warmer-than-present conditions experienced by the high latitudes of the northern hemisphere. New simulations for 6 kyr with two atmospheric/mixed-layer ocean models (Community Climate Model, version 1, CCMl, and Global ENvironmental and Ecological Simulation of Interactive Systems, version 2, GENESIS 2) are presented here and compared with results from two previous simulations with GENESIS 1 that were obtained with and without the albedo feedback due to climate-induced poleward expansion of the boreal forest. The climate model results are summarized in the form of potential vegetation maps obtained with the global BIOME model, which facilitates visual comparisons both among models and with pollen and plant macrofossil data recording shifts of the forest-tundra boundary. A preliminary synthesis shows that the forest limit was shifted 100–200 km north in most sectors. Both CCMl and GENESIS 2 produced a shift of this magnitude. GENESIS 1 however produced too small a shift, except when the boreal forest albedo feedback was included. The feedback in this case was estimated to have amplified forest expansion by approximately 50%. The forest limit changes also show meridional patterns (greatest expansion in central Siberia and little or none in Alaska and Labrador) which have yet to be reproduced by models. Further progress in understanding of the processes involved in the response of climate and vegetation to orbital forcing will require both the deployment of coupled atmosphere-biosphere-ocean models and the development of more comprehensive observational data sets

Evaluation of a Lagrangian box model using field measurements from EASE (Eastern Atlantic Summer Experiment) 1996

The Cambridge Tropospheric Trajectory model of Chemistry and Transport (CiTTyCAT), a Lagrangian chemistry model, has been evaluated using atmospheric chemical measurements collected during the East Atlantic Summer Experiment 1996 (EASE '96). This field campaign was part of the UK Natural Environment Research Council's (NERC) Atmospheric Chemistry Studies in the Oceanic Environment (ACSOE) programme, conducted at Mace Head, Republic of Ireland, during July and August 1996. The model includes a description of gas-phase tropospheric chemistry, and simple parameterisations for surface deposition, mixing from the free troposphere and emissions. The model generally compares well with the measurements and is used to study the production and loss of O3 under a variety of conditions. The mean difference between the hourly O3 concentrations calculated by the model and those measured is 0.6 ppbv with a standard deviation of 8.7 ppbv. Three specific air-flow regimes were identified during the campaign – westerly, anticyclonic (easterly) and south westerly. The westerly flow is typical of background conditions for Mace Head. However, on some occasions there was evidence of long-range transport of pollutants from North America. In periods of anticyclonic flow, air parcels had collected emissions of NOx and VOCs immediately before arriving at Mace Head, leading to O3 production. The level of calculated O3 depends critically on the precise details of the trajectory, and hence on the emissions into the air parcel. In several periods of south westerly flow, low concentrations of O3 were measured which were consistent with deposition and photochemical destruction inside the tropical marine boundary layer.

Stimulation of phosphatidylinositol hydrolysis, protein kinase C translocation, and mitogen-activated protein kinase activity by bradykinin in rat ventricular myocytes: dissociation from the hypertrophic response.

In ventricular myocytes cultured from neonatal rat hearts, bradykinin (BK), kallidin or BK(1-8) [(Des-Arg9)BK] stimulated PtdinsP2 hydrolysis by 3-4-fold. EC50 values were 6 nM (BK), 2 nM (kallidin), and 14 microM [BK(1-8)]. BK or kallidin stimulated the rapid (less than 30 s) translocation of more than 80% of the novel protein kinase C (PKC) isoforms nPKC-delta and nPKC-epsilon from the soluble to the particulate fraction. EC50 values for nPKC-delta translocation by BK or kallidin were 10 and 2 nM respectively. EC50 values for nPKC-epsilon translocation by BK or kallidin were 2 and 0.6 nM respectively. EC50 values for the translocation of nPKC-delta and nPKC-epsilon by BK(1-8) were more than 5 microM. The classical PKC, cPKC-alpha, and the atypical PKC, nPKC-zeta, did not translocate. BK caused activation and phosphorylation of p42-mitogen-activated protein kinase (MAPK) (maximal at 3-5 min, 30-35% of p42-MAPK phosphorylated). p44-MAPK was similarly activated. EC50 values for p42/p44-MAPK activation by BK were less than 1 nM whereas values for BK(1-8) were more than 10 microM. The order of potency [BK approximately equal to kallidin > BK (1-8)] for the stimulation of PtdInsP2 hydrolysis, nPKC-delta and nPKC-epsilon translocation, and p42/p44-MAPK activities suggests involvement of the B2 BK receptor subtype. In addition, stimulation of all three processes by BK was inhibited by the B2BK receptor-selective antagonist HOE140 but not by the B1-selective antagonist Leu8BK(1-8). Exposure of cells to phorbol 12-myristate 13-acetate for 24 h inhibited subsequent activation of p42/p44-MAPK by BK suggesting participation of nPKC (and possibly cPKC) isoforms in the activation process. Thus, like hypertrophic agents such as endothelin-1 (ET-1) and phenylephrine (PE), BK activates PtdInsP2 hydrolysis, translocates nPKC-delta, and nPKC-epsilon, and activates p42/p44-MAPK. However, in comparison with ET-1 and PE, BK was only weakly hypertrophic as assessed by cell morphology and patterns of gene expression. This difference could not be attributed to dissimilarities between the duration of activation of p42/p44-MAPK by BK or ET-1. Thus activation of these signalling pathways alone may be insufficient to induce a powerful hypertrophic response.