Algoritmos inteligentes para auto-correção, auto-verificação e auto-validação de medidas em sensores

This work presents a set of intelligent algorithms with the purpose of correcting calibration errors in sensors and reducting the periodicity of their calibrations. Such algorithms were designed using Artificial Neural Networks due to its great capacity of learning, adaptation and function approximation. Two approaches willbe shown, the firstone uses Multilayer Perceptron Networks to approximate the many shapes of the calibration curve of a sensor which discalibrates in different time points. This approach requires the knowledge of the sensor s functioning time, but this information is not always available. To overcome this need, another approach using Recurrent Neural Networks was proposed. The Recurrent Neural Networks have a great capacity of learning the dynamics of a system to which it was trained, so they can learn the dynamics of a sensor s discalibration. Knowingthe sensor s functioning time or its discalibration dynamics, it is possible to determine how much a sensor is discalibrated and correct its measured value, providing then, a more exact measurement. The algorithms proposed in this work can be implemented in a Foundation Fieldbus industrial network environment, which has a good capacity of device programming through its function blocks, making it possible to have them applied to the measurement process

Redes neurais artificiais no ambiente de redes industriais foundation fieldbus usando blocos funcionais padrões

Industrial automation networks is in focus and is gradually replacing older architectures of systems used in automation world. Among existing automation networks, most prominent standard is the Foundation Fieldbus (FF). This particular standard was chosen for the development of this work thanks to its complete application layer specification and its user interface, organized as function blocks and that allows interoperability among different vendors' devices. Nowadays, one of most seeked solutions on industrial automation are the indirect measurements, that consist in infering a value from measures of other sensors. This can be made through implementation of the so-called software sensors. One of the most used tools in this project and in sensor implementation are artificial neural networks. The absence of a standard solution to implement neural networks in FF environment makes impossible the development of a field-indirect-measurement project, besides other projects involving neural networks, unless a closed proprietary solution is used, which dos not guarantee interoperability among network devices, specially if those are from different vendors. In order to keep the interoperability, this work's goal is develop a solution that implements artificial neural networks in Foundation Fieldbus industrial network environment, based on standard function blocks. Along the work, some results of the solution's implementation are also presented

Estratégia de escalonamento de controladores PID baseado em regras Fuzzy para redes industriais foundation fieldbus usando blocos padrões

The main objective of work is to show procedures to implement intelligent control strategies. This strategies are based on fuzzy scheduling of PID controllers, by using only standard function blocks of this technology. Then, the standardization of Foundation Fieldbus is kept. It was developed an environment to do the necessary tests, it validates the propose. This environment is hybrid, it has a real module (the fieldbus) and a simulated module (the process), although the control signals and measurement are real. Then, it is possible to develop controllers projects. In this work, a fuzzy supervisor was developed to schedule a network of PID controller for a non-linear plant. Analyzing its performance results to the control and regulation problem

Projeto e Implementação em Ambiente Foundation Fieldbus de Filtragem Estocástica Baseada em Análise de Componentes Independentes (M175)

This work considers the development of a filtering system composed of an intelligent algorithm, that separates information and noise coming from sensors interconnected by Foundation Fieldbus (FF) network. The algorithm implementation will be made through FF standard function blocks, with on-line training through OPC (OLE for Process Control), and embedded technology in a DSP (Digital Signal Processor) that interacts with the fieldbus devices. The technique ICA (Independent Component Analysis), that explores the possibility of separating mixed signals based on the fact that they are statistically independent, was chosen to this Blind Source Separation (BSS) process. The algorithm and its implementations will be Presented, as well as the results

Viral mediated expression of insulin-like growth factor I blocks the aging-related loss of skeletal muscle function

During the aging process, mammals lose up to a third of their skeletal muscle mass and strength. Although the mechanisms underlying this loss are not entirely understood, we attempted to moderate the loss by increasing the regenerative capacity of muscle. This involved the injection of a recombinant adeno-associated virus directing overexpression of insulin-like growth factor I (IGF-I) in differentiated muscle fibers. We demonstrate that the IGF-I expression promotes an average increase of 15% in muscle mass and a 14% increase in strength in young adult mice, and remarkably, prevents aging-related muscle changes in old adult mice, resulting in a 27% increase in strength as compared with uninjected old muscles. Muscle mass and fiber type distributions were maintained at levels similar to those in young adults. We propose that these effects are primarily due to stimulation of muscle regeneration via the activation of satellite cells by IGF-I. This supports the hypothesis that the primary cause of aging-related impairment of muscle function is a cumulative failure to repair damage sustained during muscle utilization. Our results suggest that gene transfer of IGF-I into muscle could form the basis of a human gene therapy for preventing the loss of muscle function associated with aging and may be of benefit in diseases where the rate of damage to skeletal muscle is accelerated.

A biologic function for an "orphan" messenger: D-myo-inositol 3,4,5,6-tetrakisphosphate selectively blocks epithelial calcium-activated chloride channels.

Inositol phosphates are a family of water-soluble intracellular signaling molecules derived from membrane inositol phospholipids. They undergo a variety of complex interconversion pathways, and their levels are dynamically regulated within the cytosol in response to a variety of agonists. Relatively little is known about the biological function of most members of this family, with the exception of inositol 1,4,5-trisphosphate. Specifically, the biological functions of inositol tetrakisphosphates are largely obscure. In this paper, we report that D-myo-inositol 3,4,5,6-tetrakisphosphate (D-Ins(3,4,5,6)P4) has a direct biphasic (activation/inhibition) effect on an epithelial Ca(2+)-activated chloride channel. The effect of D-Ins(3,4,5,6)P4 is not mimicked by other inositol tetrakisphosphate isomers, is dependent on the prevailing calcium concentration, and is influenced when channels are phosphorylated by calmodulin kinase II. The predominant effect of D-Ins(3,4,5,6)P4 on phosphorylated channels is inhibitory at levels of intracellular calcium observed in stimulated cells. Our findings indicate the biological function of a molecule hitherto considered as an "orphan" messenger. They suggest that the molecular target for D-Ins(3,4,5,6)P4 is a plasma membrane Ca(2+)-activated chloride channel. Regulation of this channel by D-Ins(3,4,5,6)P4 and Ca2+ may have therapeutic implications for the disease states of both diabetic nephropathy and cystic fibrosis.

The effects of muscle fatigue on knee function during landing

The human knee acts as a sophisticated shock absorber during landing movements. The ability of the knee to perform this function in the real world is remarkable given that the context of the landing movement may vary widely between performances. For this reason, humans must be capable of rapidly adjusting the mechanical properties of the knee under impact load in order to satisfy many competing demands. However, the processes involved in regulating these properties in response to changing constraints remain poorly understood. In particular, the effects of muscle fatigue on knee function during step landing are yet to be fully explored. Fatigue of the knee muscles is significant for 2 reasons. First, it is thought to have detrimental effects on the ability of the knee to act as a shock absorber and is considered a risk factor for knee injury. Second, fatigue of knee muscles provides a unique opportunity to examine the mechanisms by which healthy individuals alter knee function. A review of the literature revealed that the effect of fatigue on knee function during landing has been assessed by comparing pre and postfatigue measurements, with fatigue induced by a voluntary exercise protocol. The information is limited by inconsistent results with key measures, such as knee stiffness, showing varying results following fatigue, including increased stiffness, decreased stiffness or failure to detect any change in some experiments. Further consideration of the literature questions the validity of the models used to induce and measure fatigue, as well as the pre-post study design, which may explain the lack of consensus in the results. These limitations cast doubt on the usefulness of the available information and identify a need to investigate alternative approaches. Based on the results of this review, the aims of this thesis were to: • evaluate the methodological procedures used in validation of a fatigue model • investigate the adaptation and regulation of post-impact knee mechanics during repeated step landings • use this new information to test the effects of fatigue on knee function during a step-landing task. To address the aims of the thesis, 3 related experiments were conducted that collected kinetic, kinematic and electromyographic data from 3 separate samples of healthy male participants. The methodologies involved optoelectronic motion capture (VICON), isokinetic dynamometry (System3 Pro, BIODEX) and wireless surface electromyography (Zerowire, Aurion, Italy). Fatigue indicators and knee function measures used in each experiment were derived from the data. Study 1 compared the validity and reliability of repetitive stepping and isokinetic contractions with respect to fatigue of the quadriceps and hamstrings. Fifteen participants performed 50 repetitions of each exercise twice in randomised order, over 4 sessions. Sessions were separated by a minimum of 1 week’s rest, to ensure full recovery. Validity and reliability depended on a complex interaction between the exercise protocol, the fatigue indicator, the individual and the muscle of interest. Nevertheless, differences between exercise protocols indicated that stepping was less effective in eliciting valid and reliable changes in peak power and spectral compression, compared with isokinetic exercise. A key finding was that fatigue progressed in a biphasic pattern during both exercises. The point separating the 2 phases, known as the transition point, demonstrated superior between-test reliability during the isokinetic protocol, compared with stepping. However, a correction factor should be used to accurately apply this technique to the study of fatigue during landing. Study 2 examined alterations in knee function during repeated landings, with a different sample (N =12) performing 60 consecutive step landing trials. Each landing trial was separated by 1-minute rest periods. The results provided new information in relation to the pre-post study design in the context of detecting adjustments in knee function during landing. First, participants significantly increased or decreased pre-impact muscle activity or post-impact mechanics despite environmental and task constraints remaining unchanged. This is the 1st study to demonstrate this effect in healthy individuals without external feedback on performance. Second, single-subject analysis was more effective in detecting alterations in knee function compared to group-level analysis. Finally, repeated landing trials did not reduce inter-trial variability of knee function in some participants, contrary to assumptions underpinning previous studies. The results of studies 1 and 2 were used to modify the design of Study 3 relative to previous research. These alterations included a modified isokinetic fatigue protocol, multiple pre-fatigue measurements and singlesubject analysis to detect fatigue-related changes in knee function. The study design incorporated new analytical approaches to investigate fatiguerelated alterations in knee function during landing. Participants (N = 16) were measured during multiple pre-fatigue baseline trial blocks prior to the fatigue model. A final block of landing trials was recorded once the participant met the operational fatigue definition that was identified in Study 1. The analysis revealed that the effects of fatigue in this context are heavily dependent on the compensatory response of the individual. A continuum of responses was observed within the sample for each knee function measure. Overall, preimpact preparation and post-impact mechanics of the knee were altered with highly individualised patterns. Moreover, participants used a range of active or passive pre-impact strategies to adapt post-impact mechanics in response to quadriceps fatigue. The unique patterns identified in the data represented an optimisation of knee function based on priorities of the individual. The findings of these studies explain the lack of consensus within the literature regarding the effects of fatigue on knee function during landing. First, functional fatigue protocols lack validity in inducing fatigue-related changes in mechanical output and spectral compression of surface electromyography (sEMG) signals, compared with isokinetic exercise. Second, fatigue-related changes in knee function during landing are confounded by inter-individual variation, which limits the sensitivity of group-level analysis. By addressing these limitations, the 3rd study demonstrated the efficacies of new experimental and analytical approaches to observe fatigue-related alterations in knee function during landing. Consequently, this thesis provides new perspectives into the effects of fatigue in knee function during landing. In conclusion: • The effects of fatigue on knee function during landing depend on the response of the individual, with considerable variation present between study participants, despite similar physical characteristics. • In healthy males, adaptation of pre-impact muscle activity and postimpact knee mechanics is unique to the individual and reflects their own optimisation of demands such as energy expenditure, joint stability, sensory information and loading of knee structures. • The results of these studies should guide future exploration of adaptations in knee function to fatigue. However, research in this area should continue with reduced emphasis on the directional response of the population and a greater focus on individual adaptations of knee function.

Targeting ASCT2-mediated glutamine uptake blocks prostate cancer growth and tumour development

Glutamine is conditionally essential in cancer cells, being utilized as a carbon and nitrogen source for macromolecule production, as well as for anaplerotic reactions fuelling the tricarboxylic acid (TCA) cycle. In this study, we demonstrated that the glutamine transporter ASCT2 (SLC1A5) is highly expressed in prostate cancer patient samples. Using LNCaP and PC-3 prostate cancer cell lines, we showed that chemical or shRNA-mediated inhibition of ASCT2 function in vitro decreases glutamine uptake, cell cycle progression through E2F transcription factors, mTORC1 pathway activation and cell growth. Chemical inhibition also reduces basal oxygen consumption and fatty acid synthesis, showing that downstream metabolic function is reliant on ASCT2-mediated glutamine uptake. Furthermore, shRNA knockdown of ASCT2 in PC-3 cell xenografts significantly inhibits tumour growth and metastasis in vivo, associated with the down-regulation of E2F cell cycle pathway proteins. In conclusion, ASCT2-mediated glutamine uptake is essential for multiple pathways regulating the cell cycle and cell growth, and is therefore a putative therapeutic target in prostate cancer.

The insecticidal spider toxin SFI1 is a knottin peptide that blocks the pore of insect voltage-gated sodium channels via a large β-hairpin loop

Spider venoms contain a plethora of insecticidal peptides that act on neuronal ion channels and receptors. Because of their high specificity, potency and stability, these peptides have attracted much attention as potential environmentally friendly insecticides. Although many insecticidal spider venom peptides have been isolated, the molecular target, mode of action and structure of only a small minority have been explored. Sf1a, a 46-residue peptide isolated from the venom of the tube-web spider Segesteria florentina, is insecticidal to a wide range of insects, but nontoxic to vertebrates. In order to investigate its structure and mode of action, we developed an efficient bacterial expression system for the production of Sf1a. We determined a high-resolution solution structure of Sf1a using multidimensional 3D/4D NMR spectroscopy. This revealed that Sf1a is a knottin peptide with an unusually large β-hairpin loop that accounts for a third of the peptide length. This loop is delimited by a fourth disulfide bond that is not commonly found in knottin peptides. We showed, through mutagenesis, that this large loop is functionally critical for insecticidal activity. Sf1a was further shown to be a selective inhibitor of insect voltage-gated sodium channels, consistent with its 'depressant' paralytic phenotype in insects. However, in contrast to the majority of spider-derived sodium channel toxins that function as gating modifiers via interaction with one or more of the voltage-sensor domains, Sf1a appears to act as a pore blocker.

Ternary function and circuit design using ternary multiplexers

The recent trend towards minimizing the interconnections in large scale integration (LSI) circuits has led to intensive investigation in the development of ternary circuits and the improvement of their design. The ternary multiplexer is a convenient and useful logic module which can be used as a basic building block in the design of a ternary system. This paper discusses a systematic procedure for the simplification and realization of ternary functions using ternary multiplexers as building blocks. Both single level and multilevel multiplexing techniques are considered. The importance of the design procedure is highlighted by considering two specific applications, namely, the development of ternary adder/subtractor and TCD to ternary converter.

Structure and function of GDNF receptor alpha splice variants

The growth factors of the glial cell line-derived neurotrophic factor (GDNF) family consisting of GDNF, neurturin (NRTN), artemin (ARTN) and persephin (PSPN), are involved in the development, differentiation and maintenance of many types of neurons. They also have important functions outside the nervous system in the development of kidney, testis and thyroid gland. Each of these GFLs preferentially binds to one of the glycosylphosphatidylinositol (GPI)-anchored GDNF family receptors α (GFRα). GDNF binds to GFRα1, NRTN to GFRα2, ARTN to GFRα3 and PSPN to GFRα4. The GFLs in the complex with their cognate GFRα receptors all bind to and signal through the receptor tyrosine kinase RET. Alternative splicing of the mouse GFRα4 gene yields three splice isoforms. These had been described as putative GPI-anchored, transmembrane and soluble forms. My goal was to characterise the function of the different forms of mouse GFRα4. I firstly found that the putative GPI-anchored GFRα4 (GFRα4-GPI) is glycosylated, membrane-bound, GPI-anchored and interacts with PSPN and RET. We also showed that mouse GFRα4-GPI mediates PSPN-induced phosphorylation of RET, promotes PSPN-dependent neuronal differentiation of the rat pheochromocytoma cell line PC6-3 and PSPN-dependent survival of cerebellar granule neurons (CGN). However, although this receptor can mediate PSPN-signalling and activate RET, GFRα4-GPI does not recruit RET into lipid rafts. The recruitment of RET into lipid rafts has previously been thought to be a crucial event for GDNF- and GFL-mediated signalling via RET. I secondly demonstrated that the putative transmembrane GFRα4 (GFRα4-TM) is indeed a real transmembrane GFRα4 protein. Although it has a weak binding capacity for PSPN, it can not mediate PSPN-dependent phosphorylation of RET, neuronal differentiation or survival. These data show that GFRα4-TM is inactive as a receptor for PSPN. Surprisingly, GFRα4-TM can negatively regulate PSPN-mediated signalling via GFRα4-GPI. GFRα4-TM interacts with GFRα4-GPI and blocks PSPN-induced phosphorylation of RET, neuronal differentiation as well as survival. Taken together, our data show that GFRα4-TM may act as a dominant negative inhibitor of PSPN-mediated signaling. The most exciting part of my work was the finding that the putative soluble GFRα4 (GFRα4-sol) can form homodimers and function as an agonist of the RET receptor. In the absence of PSPN, GFRα4-sol can promote the phosphorylation of RET, trigger the activation of the PI-3K/AKT pathway, induce neuronal differentiation and support the survival of CGN. Our findings are in line with a recent publication showing the GFRα4-sol might contribute to the inherited cancer syndrome multiple endocrine neoplasia type 2. Our data provide an explanation to how GFRα4-sol may cause or modify the disease. Mammalian GFRα4 receptors all lack the first Cys-rich domain which is present in other GFRα receptors. In the final part of my work I have studied the function of this particular domain. I created a truncated GFRα1 construct lacking the first Cys-rich domain. Using binding assays in both cellular and cell-free systems, phosphorylation assays with RET, as well as neurite outgrowth assays, we found that the first Cys-rich domain contributes to an optimal function of GFRα1, by stabilizing the interaction between GDNF and GFRα1.

Asymmetric cross-coupled differential pair configuration to realize neuron activation function and its derivative

In this brief, we present a new circuit technique to generate the sigmoid neuron activation function (NAF) and its derivative (DNAF). The circuit makes use of transistor asymmetry in cross-coupled differential pair to obtain the derivative. The asymmetry is introduced through external control signal, as and when required. This results in the efficient utilization of the hard-ware by realizing NAF and DNAF using the same building blocks. The operation of the circuit is presented in the subthreshold region for ultra low-power applications. The proposed circuit has been experimentally prototyped and characterized as a proof of concept on the 1.5-mum AMI technology.

Rocking response of rectangular rigid blocks under random noise base excitations

The response of a rigid rectangular block resting on a rigid foundation and acted upon simultaneously by a horizontal and a vertical random white-noise excitation is considered. In the equation of motion, the energy dissipation is modeled through a viscous damping term. Under the assumption that the body does not topple, the steady-state joint probability density function of the rotation and the rotational velocity is obtained using the Fokker-Planck equation approach. Closed form solution is obtained for a specific combination of system parameters. A more general but approximate solution to the joint probability density function based on the method of equivalent non-linearization is also presented. Further, the problem of overturning of the block is approached in the framework of the diffusion methods for first passage failure studies. The overturning of the block is deemed incipient when the response trajectories in the phase plane cross the separatrix of the conservative unforced system. Expressions for the moments of first passage time are obtained via a series solution to the governing generalized Pontriagin-Vitt equations. Numerical results illustra- tive of the theoretical solutions are presented and their validity is examined through limited amount of digital simulations.

Lossless and lossy image compression using Boolean function minimization

A novel approach for lossless as well as lossy compression of monochrome images using Boolean minimization is proposed. The image is split into bit planes. Each bit plane is divided into windows or blocks of variable size. Each block is transformed into a Boolean switching function in cubical form, treating the pixel values as output of the function. Compression is performed by minimizing these switching functions using ESPRESSO, a cube based two level function minimizer. The minimized cubes are encoded using a code set which satisfies the prefix property. Our technique of lossless compression involves linear prediction as a preprocessing step and has compression ratio comparable to that of JPEG lossless compression technique. Our lossy compression technique involves reducing the number of bit planes as a preprocessing step which incurs minimal loss in the information of the image. The bit planes that remain after preprocessing are compressed using our lossless compression technique based on Boolean minimization. Qualitatively one cannot visually distinguish between the original image and the lossy image and the value of mean square error is kept low. For mean square error value close to that of JPEG lossy compression technique, our method gives better compression ratio. The compression scheme is relatively slower while the decompression time is comparable to that of JPEG.