No description available.

One of the key determinants of an organization’s success is its ability to adapt to marketplace change. Given this reality, how do organizations survive or even thrive in today’s dynamic markets? The answer to this question is highly related to the adaptability of one of the organization’s key resource: its employees. Indeed, the central component of an organization’s success will depend on its ability to drive changes in the mental models of individual employees. Moreover, a critical facilitator of that will be the development of decision support tools that support change of those mental models. In response to this need there has been a tremendous growth in business analytic decision support tools, estimated to reach almost $200 billion in sales by 2019. The premise of this research is that these decision support tools are ill-suited to support true mental model change because they have focused on a feedback-enabled view and generally lack a predictive (feedforward-enabled) view of the likely outcomes of the decision. The purpose of this research is to study how changes in mental models can be facilitated through this feedforward mechanisms within the DSS tool. This research used a mixed method approach, leveraging the strengths of quantitative and qualitative research methodologies, to study this research question. The research showed that the feedforward-enabled DSS tool did create more mental model change and alignment (versus an ideal solution) compared to the control. The feedforward enabled tool also produced better alignment than the feedback-enabled decision support tool. In fact, the feedback-enabled decision support was shown to result in a poorer alignment with the ideal solution. This paper concludes by suggesting five areas for future research. / Business Administration/Management Information Systems

Over the past 20 years research into miniature actuators has been increasing. In addition to having a compact geometry, desirable characteristics for miniature actuators include having a large power-to-weight ratio, fast response, fine resolution of movement and high power efficiency. In the first part of this thesis the design of a miniature rotary robot joint is presented. Two single acting miniature cylinders each with a bore diameter of 4 mm drive the joint using water as the hydraulic fluid. The cylinders are mated to a rack and pinion mechanism that converts the opposing linear motion of the cylinders shafts into rotation. Also within the design, a novel position sensor using magnetic field sensing technology is presented. Overall, the joint measures 11 mm wide x 8.8 mm high x 150 mm long. In the second part of this thesis a hydraulic servo positioning system is presented along with a novel valve modeling technique and two position control strategies. Four low-cost, 3-way on/off solenoid valves were used to control the flow of the water in and out of the cylinders. The two nonlinear position controllers employed were a position-velocity-acceleration plus model-based feedforward controller (PVA+FF) and a novel PVA + FF plus sliding mode controller. For experiments involving horizontal rotation of the joint while carrying no load the PVA +FF controller achieved a steady-state error of ± 0.77° or ± 0.06 mm in terms of rack position. The steady-state error produced by the PVA + FF plus sliding mode controller was ± 0.85° or ± 0.07 mm. The maximum tracking error produced by both controllers was 5° or 0.41 mm and occurred during the initial cycloidal rising portion of a 120° displacement. The root mean square error (RMSE) of the PVA + FF and PVA + FF plus sliding mode controllers were 42% and 54% less than that produced by a linear PVA controller. Both controllers were found to be robust to changes in payload. This was experimentally verified by adding masses of 6.5 g and 13.5 g to the end of the output link of the joint. By conducting similar experiments in the vertical direction it was found that the PVA + FF plus sliding mode controller was more robust, achieving on average a 30% reduction in RMSE compared to the FF + PVA controller. / Thesis / Master of Applied Science (MASc)

This dissertation presents a set of general numerical tools for simulating feedforward active noise control in the frequency domain. Feedforward control is numerically similar to linear least squares regression, and can take advantage of various numerical techniques developed in the statistics literature for use with regression. Therefore, an important theme of this work is to look at the control problem from a statistical point of view, and explore the analogies between feedforward control and basic statistical principles of regression. Motivating the numerical approach is the need to simulate active noise control for systems whose dynamics must be modeled numerically because analytical solutions do not exist, e.g., fluid-structure interaction problems. Plant dynamics for examples in the present work are modeled using a finite-element / boundary-element computer program, and the associated numerical methods are general enough for us with many types of problems. The derivation is presented in the context of active structural-acoustic control (ASAC), in which sound radiating from a vibrating structure is controlled by applying time-harmonic vibrational inputs directly on the structure. First, a feedforward control simulation is developed for a submerged spherical shell using both analytical and numerical techniques; the numerical formulation is found by discretizing the integrations used in the analytical approach. ASAC is shown to be effective for controlling radiation from the spherical shell. For a point-force disturbance at low frequencies, a single control input can reduce the radiated power by up to 20 dB (ignoring the possibility of measurement noise). A more general numerical methodology is then developed based on weighted least-squares regression in the complex domain. It is shown that basic regression diagnostics, which are used in the statistics literature to describe the quality and reliability of a regression, can be used to model the effects of error sensor measurement noise to produce a more realistic simulation. Numerical results are presented for a finite-length, fluid-loaded cylindrical shell with clamped, rigid end closures. It is shown that when the controller reduces the radiated power by less than 2 dB, the control simulation is usually invalid for statistical reasons. Also developed are confidence intervals for the individual control input magnitudes, and prediction intervals which help evaluate the sensitivity to measurement noise for the regression as a whole. Collinearity, a type of numerical ill-conditioning that can corrupt regression results, is demonstrated to occur in an example feedforward control simulation. The effects of collinearity are discussed, and a basic diagnostic is developed to detect and analyze collinearity. Subset selection, a numerical procedure for improving regressions, is shown to correspond to optimizing actuator locations for best control system performance. Exhaustive-search subset selection is used to optimize actuator locations for a sample structure. Finally, a convenient method is given for investigating alternate controller formulations, and examples of several alternate controllers are given including a wavenumber-domain controller. Numerical results for a cylindrical shell give insight to the mechanisms used by the control system, and a new visualization technique is used to relate farfield pressure distributions to surface velocity distributions using wavenumber analysis. / Ph. D.

The concept of using heat flux measurements to provide the input for a feedforward temperature control loop is investigated. The feedforward loop is added to proportional and integral feedback control to increase the speed of the response to a disturbance. Comparison is made between the feedback and the feedback plus feedforward control laws. The control law with the feedforward control loop is also compared to the conventional approach of adding derivative control to speed up the system response to a disturbance. The concept was tested using a simple flat plate heated on one side and exposed to a step change in the convective heat loss on the other side. A controller was constructed using an analog computer to compare the feedforward and feedback approaches. The conventional control approach was tested using a commercial temperature controller. The feedback and feedforward approaches were also simulated. The results showed that the feedforward control approach produced significant improvements in the response to the disturbance. The integral of the squared error between the setpoint and actual temperature was reduced by approximately 90 percent by the addition of feedforward control to the feedback control. The maximum temperature deviation from the setpoint was also reduced by 70 percent with the addition of feedforward control. Qualitative agreement was obtained between the experimental results and the computer simulations. The conventional approach of adding derivative control to the proportional and integral control showed an increase of 20 percent in the integral of the squared error, but offered no significant improvement in the maximum temperature deviation. The addition of derivative control also caused the stability of the system to decrease, while the addition of feedforward had no adverse effects on the system stability. The concept of using heat flux measurements for feedforward control was successfully demonstrated by both simulations and experiments. / Master of Science

This quasi-experimental field study examined the efficacy of a safety self-management intervention to increase safety-related work practices in mining operations. A total of 15 male miners participated in the study while engaging in their normal work practices at the Virginia Tech Quarry, located in Blacksburg, Virginia. The study had two groups, Feedforward (n=8)--participants self-recorded their intentions to engage in specific percentages of safety-related work behaviors before starting their shift for the day, and Feedback (n=7)--participants self-recorded their percentages of safety-related work behaviors after their shift for the day. After a seven-week Baseline, miners participated in a safety training presentation. Immediately following this training, participants from each group were instructed to complete one self-monitoring form each day on their self-intentions (Feedforward) or actual (Feedback) safety performance for four weeks. Participants were paid $1.00 for each completed self-monitoring form. All completed forms were entered into a raffle for a cash prize of $50.00 at the end of the Intervention phase. During Withdrawal (four weeks) miners did not complete any self-monitoring forms. Trained research assistants made a total of 10, 905 obtrusive behavioral observations on three target behaviors (ear plugs, dust mask, and safety glasses) and five non-target behaviors (gloves, hard hat, boots, knee position during lifts, body position during lifts) across phases. Results showed the safety self-management intervention significantly increased safety performance across both target and non-target behaviors during the Intervention phase. / Master of Science

Traditional active noise control (ANC) systems rely on a physical sensor to measure the error signal at the desired location of attenuation. The error signal is then used to update an adaptive controller, which ultimately attenuates the measured response. However, it is not always practical to use traditional ANC in real-world applications. For example, as small unmanned aerial systems (UAS) become more commonly used, community noise exposure also increases, along with the desire to reduce UAS noise. Traditional ANC systems that rely on physical sensors at observer locations are impractical, since a UAS does not typically have real-time access to the response at an observer's ears, which is realistically in the far-field. Virtual error sensing (VES) can augment an ANC system using near-field measurements to estimate the response at a desired far-field location. In this way, the VES technique effectively shifts the zone of quiet from the location of the physical sensor(s) to a different "virtual" location. This thesis begins by outlining past work that used traditional ANC methods and virtual error sensing techniques. Numerical modeling results showing the predicted spatial change in SPL achieved using a virtual sensor will be presented. Experimental tests used ANC to attenuate the noise from a single UAS rotor at far-field locations using a near-field microphone and the remote microphone technique (RMT) to develop the VES. The results of the VES alone and with an ANC approach at several far-field virtual locations will be presented and discussed. / Master of Science / Small unmanned aerial systems (sUAS) are becoming increasingly common for private, military, and commercial use, and as such, community noise exposure is increasing. Reducing the noise produced by UAS could help improve community acceptance. Active noise control (ANC) might be used to attenuate noise produced by sUAS, however, traditional ANC systems would require a physical sensor in the far-field, which is not feasible. A virtual error sensor (VES) could eliminate the need for a far-field sensor. This thesis describes the proposed VES strategy, and presents numerical simulations and experimental results that highlight both the benefits and limitations of the approach. Results of the VES system with and without an ANC approach are discussed. Experimental testing focused on attenuating the tonal noise produced by one 2-bladed rotor with a tip radius of 4.7 inches. Pressure variations caused by the blade rotation were measured in the near and far-field using electret microphones and externally polarized condenser microphones, respectively. The ANC system used the filtered-x least mean squares algorithm in conjunction with the VES system to estimate the far-field response. A 2-inch diameter speaker served as the secondary source to provide the appropriate control input to the system. Experimental results show reductions between 6-13 dB at varying far-field locations and rotation rates.

This thesis focuses on data-driven methods, specifically artificial neural networks, to identify material model parameters in high density polyethylene (HDPE) for finite element (FE) simulations. The study thoroughly examines the anisotropy in HDPE by testing different material orientations with digital image correlation (DIC) during uniaxial tensile tests. DIC enabled precise measurement of strain distribution,unveiling both diffuse and local necking strain. Two hardening models,the Swift-Hockett-Sherby (S/HS) and a custom model, were explored to characterize HDPE’s plastic behaviour in FE simulations. In consistencies between predicted outcomes using the SHS model and experimental results prompt the consideration of custom equations forenhanced accuracy. The Hill48 yield model was introduced for the FE model to cover the anisotropic properties of the material. Large datasets were generated from these simulations to cover a wide range of different material configurations. The datasets were used to train neural networks so that a wide range of different HDPE grades can later be fed to the network to determine the associated material parameters. An Abaqus-Isight model was developed to automate parameter variation, simulation, and data extraction, thus streamlining the process and saving time. Data extracted from simulations, including force displacement and strain, are leveraged for neural network training. The study evaluated two types of neural networks: feed forward neural networks (FFNN) and long short-term memory neural networks(LSTM). It was found that FFNN performed better than LSTM for this task. Therefore, the research focused more on refining the FFNN approach. Overall, the implementation of the custom hardening modelin combination with the Hill48 yield model was successful, but showed weaknesses in CD orientation

The majority of previous research investigating the role of vision in controlling adaptive gait has predominantly focused on over-ground walking or obstacle negotiation. Thus there is a paucity of literature investigating visuomotor control of step descent. This thesis addressed the importance of the lower visual field (lvf) in regulating step descent landing control, and determined when visual feedback is typically used in regulating landing control prior to / during step descent. When step descents were completed from a stationary starting position, with the lvf occluded or degraded, participants adapted their stepping strategy in a manner consistent with being uncertain regarding the precise location of the foot / lower leg relative to the floor. However, these changes in landing control under conditions of lvf occlusion were made without fundamentally altering stepping strategy. This suggests that participants were able to plan the general stepping strategy when only upper visual field cues were available. When lvf was occluded from either 2 or 1 step(s) prior to descending a step during on-going gait, stepping strategy was only affected when the lvf was occluded in the penultimate step. Findings suggest that lvf cues are acquired in the penultimate step / few seconds prior to descent and provide exproprioceptive information of the foot / lower leg relative to the floor which ensures landing is regulated with increased certainty. Findings also highlight the subtle role of online vision used in the latter portion of step descent to 'fine tune' landing control.

Considerando a importÃncia do controle da coagulaÃÃo quÃmica para o processo de tratamento de Ãgua por filtraÃÃo rÃpida, esta dissertaÃÃo propÃe a aplicaÃÃo de redes neurais artificiais para inferÃncia dos sinais de controle de dosagem de coagulantes principal e auxiliar, no processo de coagulaÃÃo quÃmica em uma estaÃÃo de tratamento de Ãgua por filtraÃÃo rÃpida. Para tanto, foi feito uma anÃlise comparativa da aplicaÃÃo de modelos baseados em redes neurais do tipo: alimentada adiante focada atrasada no tempo (FTLFN); alimentada adiante atrasada no tempo distribuÃda (DTLFN); recorrente de Elman (ERN) e auto-regressiva nÃo-linear com entradas exÃgenas (NARX). Da anÃlise comparativa, o modelo baseado em redes NARX apresentou melhores resultados, evidenciando o potencial do modelo para uso em casos reais, o que contribuirà para a viabilizaÃÃo de projetos desta natureza em estaÃÃes de tratamento de Ãgua de pequeno porte. / Considering the importance of the chemical coagulation control for the water treatment by direct filtration, this work proposes the application of artificial neural networks for inference of dosage control signals of principal and auxiliary coagulant, in the chemical coagulation process in a water treatment plant by direct filtration. To that end, was made a comparative analysis of the application of models based on neural networks, such as: Focused Time Lagged Feedforward Network (FTLFN); Distributed Time Lagged Feedforward Network (DTLFN); Elman Recurrent Network (ERN) and Non-linear Autoregressive with exogenous inputs (NARX). From the comparative analysis, the model based on NARX networks showed better results, demonstrating the potential of the model for use in real cases, which will contribute to the viability of projects of this nature in small size water treatment plants.