Dynamic recurrent neural network for system identification and control

A dynamic recurrent neural network (DRNN) that can be viewed as a generalisation of the Hopfield neural network is proposed to identify and control a class of control affine systems. In this approach, the identified network is used in the context of the differential geometric control to synthesise a state feedback that cancels the nonlinear terms of the plant yielding a linear plant which can then be controlled using a standard PID controller.

A mixture of experts network structure construction algorithm for modelling and control

This paper introduces a new fast, effective and practical model structure construction algorithm for a mixture of experts network system utilising only process data. The algorithm is based on a novel forward constrained regression procedure. Given a full set of the experts as potential model bases, the structure construction algorithm, formed on the forward constrained regression procedure, selects the most significant model base one by one so as to minimise the overall system approximation error at each iteration, while the gate parameters in the mixture of experts network system are accordingly adjusted so as to satisfy the convex constraints required in the derivation of the forward constrained regression procedure. The procedure continues until a proper system model is constructed that utilises some or all of the experts. A pruning algorithm of the consequent mixture of experts network system is also derived to generate an overall parsimonious construction algorithm. Numerical examples are provided to demonstrate the effectiveness of the new algorithms. The mixture of experts network framework can be applied to a wide variety of applications ranging from multiple model controller synthesis to multi-sensor data fusion.

Neural network control of a simple mobile robot

In recent years researchers in the Department of Cybernetics have been developing simple mobile robots capable of exploring their environment on the basis of the information obtained from a few simple sensors. These robots are used as the test bed for exploring various behaviours of single and multiple organisms: the work is inspired by considerations of natural systems. In this paper we concentrate on that part of the work which involves neural networks and related techniques. These neural networks are used both to process the sensor information and to develop the strategy used to control the robot. Here the robots, their sensors, and the neural networks used and all described. 1.

The system identification and control of Hammerstein system using non-uniform rational B-spline neural network and particle swarm optimization

In this paper a new system identification algorithm is introduced for Hammerstein systems based on observational input/output data. The nonlinear static function in the Hammerstein system is modelled using a non-uniform rational B-spline (NURB) neural network. The proposed system identification algorithm for this NURB network based Hammerstein system consists of two successive stages. First the shaping parameters in NURB network are estimated using a particle swarm optimization (PSO) procedure. Then the remaining parameters are estimated by the method of the singular value decomposition (SVD). Numerical examples including a model based controller are utilized to demonstrate the efficacy of the proposed approach. The controller consists of computing the inverse of the nonlinear static function approximated by NURB network, followed by a linear pole assignment controller.

Imbalanced functional link between executive control network and reward network explain the online-game seeking behaviors in Internet gaming disorder

Literatures have shown that Internet gaming disorder (IGD) subjects show impaired executive control and enhanced reward sensitivities than healthy controls. However, how these two networks jointly affect the valuation process and drive IGD subjects' online-game-seeking behaviors remains unknown. Thirty-five IGD and 36 healthy controls underwent a resting-states scan in the MRI scanner. Functional connectivity (FC) was examined within control and reward network seeds regions, respectively. Nucleus accumbens (NAcc) was selected as the node to find the interactions between these two networks. IGD subjects show decreased FC in the executive control network and increased FC in the reward network when comparing with the healthy controls. When examining the correlations between the NAcc and the executive control/reward networks, the link between the NAcc - executive control network is negatively related with the link between NAcc - reward network. The changes (decrease/increase) in IGD subjects' brain synchrony in control/reward networks suggest the inefficient/overly processing within neural circuitry underlying these processes. The inverse proportion between control network and reward network in IGD suggest that impairments in executive control lead to inefficient inhibition of enhanced cravings to excessive online game playing. This might shed light on the mechanistic understanding of IGD.

Real-Time software control of neural stimulation in a closed loop Animat system

A recent area for investigation into the development of adaptable robot control is the use of living neuronal networks to control a mobile robot. The so-called Animat paradigm comprises a neuronal network (the ‘brain’) connected to an external embodiment (in this case a mobile robot), facilitating potentially robust, adaptable robot control and increased understanding of neural processes. Sensory input from the robot is provided to the neuronal network via stimulation on a number of electrodes embedded in a specialist Petri dish (Multi Electrode Array (MEA)); accurate control of this stimulation is vital. We present software tools allowing precise, near real-time control of electrical stimulation on MEAs, with fast switching between electrodes and the application of custom stimulus waveforms. These Linux-based tools are compatible with the widely used MEABench data acquisition system. Benefits include rapid stimulus modulation in response to neuronal activity (closed loop) and batch processing of stimulation protocols.

Epidemiological risk assessment using linked network and grid based modelling: Phytophthora ramorum and Phytophthora kernoviae in the UK

We developed a stochastic simulation model incorporating most processes likely to be important in the spread of Phytophthora ramorum and similar diseases across the British landscape (covering Rhododendron ponticum in woodland and nurseries, and Vaccinium myrtillus in heathland). The simulation allows for movements of diseased plants within a realistically modelled trade network and long-distance natural dispersal. A series of simulation experiments were run with the model, representing an experiment varying the epidemic pressure and linkage between natural vegetation and horticultural trade, with or without disease spread in commercial trade, and with or without inspections-with-eradication, to give a 2 x 2 x 2 x 2 factorial started at 10 arbitrary locations spread across England. Fifty replicate simulations were made at each set of parameter values. Individual epidemics varied dramatically in size due to stochastic effects throughout the model. Across a range of epidemic pressures, the size of the epidemic was 5-13 times larger when commercial movement of plants was included. A key unknown factor in the system is the area of susceptible habitat outside the nursery system. Inspections, with a probability of detection and efficiency of infected-plant removal of 80% and made at 90-day intervals, reduced the size of epidemics by about 60% across the three sectors with a density of 1% susceptible plants in broadleaf woodland and heathland. Reducing this density to 0.1% largely isolated the trade network, so that inspections reduced the final epidemic size by over 90%, and most epidemics ended without escape into nature. Even in this case, however, major wild epidemics developed in a few percent of cases. Provided the number of new introductions remains low, the current inspection policy will control most epidemics. However, as the rate of introduction increases, it can overwhelm any reasonable inspection regime, largely due to spread prior to detection. (C) 2009 Elsevier B.V. All rights reserved.

Modelling disease spread and control in networks: implications for plant sciences

Networks are ubiquitous in natural, technological and social systems. They are of increasing relevance for improved understanding and control of infectious diseases of plants, animals and humans, given the interconnectedness of today's world. Recent modelling work on disease development in complex networks shows: the relative rapidity of pathogen spread in scale-free compared with random networks, unless there is high local clustering; the theoretical absence of an epidemic threshold in scale-free networks of infinite size, which implies that diseases with low infection rates can spread in them, but the emergence of a threshold when realistic features are added to networks (e.g. finite size, household structure or deactivation of links); and the influence on epidemic dynamics of asymmetrical interactions. Models suggest that control of pathogens spreading in scale-free networks should focus on highly connected individuals rather than on mass random immunization. A growing number of empirical applications of network theory in human medicine and animal disease ecology confirm the potential of the approach, and suggest that network thinking could also benefit plant epidemiology and forest pathology, particularly in human-modified pathosystems linked by commercial transport of plant and disease propagules. Potential consequences for the study and management of plant and tree diseases are discussed.

Self-assembly of two-component gels: stoichiometric control and component selection

Two-component systems capable of self-assembling into soft gel-phase materials are of considerable interest due to their tunability and versatility. This paper investigates two-component gels based on a combination of a L-lysine-based dendron and a rigid diamine spacer (1,4-diaminobenzene or 1,4-diaminocyclohexane). The networked gelator was investigated using thermal measurements, circular dichroism, NMR spectroscopy and small angle neutron scattering (SANS) giving insight into the macroscopic properties, nanostructure and molecular-scale organisation. Surprisingly, all of these techniques confirmed that irrespective of the molar ratio of the components employed, the "solid-like" gel network always consisted of a 1:1 mixture of dendron/diamine. Additionally, the gel network was able to tolerate a significant excess of diamine in the "liquid-like" phase before being disrupted. In the light of this observation, we investigated the ability of the gel network structure to evolve from mixtures of different aromatic diamines present in excess. We found that these two-component gels assembled in a component-selective manner, with the dendron preferentially recognising 1,4-diaminobenzene (>70%). when similar competitor diamines (1,2- and 1,3-diaminobenzene) are present. Furthermore, NMR relaxation measurements demonstrated that the gel based oil 1,4-diaminobenzene was better able to form a selective ternary complex with pyrene than the gel based oil 1,4-diaminocyclohexane, indicative of controlled and selective pi-pi interactions within a three-component assembly. As such, the results ill this paper demonstrate how component selection processes in two-component gel systems call control hierarchical self-assembly.

A novel technique for differentiation of proteins in the development of acid gel structure from control and heat treated milk using confocal scanning laser microscopy

The incorporation of caseins and whey proteins into acid gels produced from unheated and heat treated skimmed milk was studied by confocal scanning laser microscopy (CSLM) using fluorescent labelled proteins. Bovine casein micelles were labelled using Alexa Fluor 594, while whey proteins were labelled using Alexa Fluor 488. Samples of the labelled protein solutions were introduced into aliquots of pasteurised skim milk, and skim milk heated to 90 degrees C for 2 min and 95 degrees C for 8 min. The milk was acidified at 40 degrees C to a final pH of 4.4 using 20 g gluconodelta-lactone/l (GDL). The formation of gels was observed with CSLM at two wavelengths (488 nm and 594 nm), and also by visual and rheological methods. In the control milk, as pH decreased distinct casein aggregates appeared, and as further pH reduction occurred, the whey proteins could be seen to coat the casein aggregates. With the heated milks, the gel structure was formed of continuous strands consisting of both casein and whey protein. The formation of the gel network was correlated with an increase in the elastic modulus for all three treatments, in relation to the severity of heat treatment. This model system allows the separate observation of the caseins and whey proteins, and the study of the interactions between the two protein fractions during the formation of the acid gel structure, on a real-time basis. The system could therefore be a valuable tool in the study of structure formation in yoghurt and other dairy protein systems.

Prefrontal social cognition network dysfunction underlying face encoding and social anxiety in fragile X syndrome

Individuals with fragile X syndrome (FXS) commonly display characteristics of social anxiety, including gaze aversion, increased time to initiate social interaction, and difficulty forming meaningful peer relationships. While neural correlates of face processing, an important component of social interaction, are altered in FXS, studies have not examined whether social anxiety in this population is related to higher cognitive processes, such as memory. This study aimed to determine whether the neural circuitry involved in face encoding was disrupted in individuals with FXS, and whether brain activity during face encoding was related to levels of social anxiety. A group of 11 individuals with FXS (5 M) and 11 age-and gender-matched control participants underwent fMRI scanning while performing a face encoding task with onlineeye-tracking. Results indicate that compared to the control group, individuals with FXS exhibited decreased activation of prefrontal regions associated with complex social cognition, including the medial and superior frontal cortex, during successful face encoding. Further, the FXS and control groups showed significantly different relationships between measures of social anxiety (including gaze-fixation) and brain activity during face encoding. These data indicate that social anxiety in FXS may be related to the inability to successfully recruit higher level social cognition regions during the initial phases of memory formation. (C) 2008 Elsevier Inc. All rights reserved.

Neural systems in the visual control of steering

Visual control of locomotion is essential for most mammals and requires coordination between perceptual processes and action systems. Previous research on the neural systems engaged by self-motion has focused on heading perception, which is only one perceptual subcomponent. For effective steering, it is necessary to perceive an appropriate future path and then bring about the required change to heading. Using function magnetic resonance imaging in humans, we reveal a role for the parietal eye fields (PEFs) in directing spatially selective processes relating to future path information. A parietal area close to PEFs appears to be specialized for processing the future path information itself. Furthermore, a separate parietal area responds to visual position error signals, which occur when steering adjustments are imprecise. A network of three areas, the cerebellum, the supplementary eye fields, and dorsal premotor cortex, was found to be involved in generating appropriate motor responses for steering adjustments. This may reflect the demands of integrating visual inputs with the output response for the control device.

Architecture for neuronal cell control of a mobile robot

It is usually expected that the intelligent controlling mechanism of a robot is a computer system. Research is however now ongoing in which biological neural networks are being cultured and trained to act as the brain of an interactive real world robot - thereby either completely replacing or operating in a cooperative fashion with a computer system. Studying such neural systems can give a distinct insight into biological neural structures and therefore such research has immediate medical implications. In particular, the use of rodent primary dissociated cultured neuronal networks for the control of mobile `animals' (artificial animals, a contraction of animal and materials) is a novel approach to discovering the computational capabilities of networks of biological neurones. A dissociated culture of this nature requires appropriate embodiment in some form, to enable appropriate development in a controlled environment within which appropriate stimuli may be received via sensory data but ultimate influence over motor actions retained. The principal aims of the present research are to assess the computational and learning capacity of dissociated cultured neuronal networks with a view to advancing network level processing of artificial neural networks. This will be approached by the creation of an artificial hybrid system (animal) involving closed loop control of a mobile robot by a dissociated culture of rat neurons. This 'closed loop' interaction with the environment through both sensing and effecting will enable investigation of its learning capacity This paper details the components of the overall animat closed loop system and reports on the evaluation of the results from the experiments being carried out with regard to robot behaviour.