- Browse by Subject
Browsing by Subject "Mechatronics"
Now showing 1 - 7 of 7
Results Per Page
Sort Options
Item Deployment of Compressed MobileNet V3 on iMX RT 1060(IEEE Xplore, 2021-04) Prasad, S. P. Kavyashree; El-Sharkawy, Mohamed; Electrical and Computer Engineering, School of Engineering and TechnologyDeep Neural Networks (DNN) are prominent in most applications today. From self-driving cars, sentiment analysis, surveillance systems, and robotics, they have been used extensively. Among DNNs, Convolutional Neural Networks (CNN) have achieved massive success in computer vision applications as the human visual system inspires their architecture. However, striving to achieve higher accuracies, CNN complexity, parameters, and layers were increased, which led to a drastic surge in their size, making their deployment challenging. Over the years, many researchers have proposed various techniques to alleviate this issue-one of them being Design Space Exploration (DSE) to minimize size and computation with little compromise to accuracy. MobileNet V3 is one such architecture designed to achieve good accuracy while being mindful of resources. It produces an accuracy of 88.93% on CIFAR-10 with a size of 15.3MB. This paper further reduces its size to 2.3MB while boosting its accuracy to 89.13% using DSE techniques. It is then deployed into NXP's i.MX RT1060 Advanced Driver Assistance System (ADAS) platform.Item Development of a Mechatronics Instrument Assisted Soft Tissue Mobilization (IASTM) Device to Quantify Force and Orientation Angles(2016-05) Alotaibi, Ahmed Mohammed; Anwar, Sohel; Loghmani, Mary T.; Chien, Stanley Yung-PingInstrument assisted soft tissue mobilization (IASTM) is a form of massage using rigid manufactured or cast devices. The delivered force, which is a critical parameter in massage during IASTM, has not been measured or standardized for most clinical practices. In addition to the force, the angle of treatment and frequency play an important role during IASTM. As a result, there is a strong need to characterize the delivered force to a patient, angle of treatment, and stroke frequency. This thesis proposes two novel mechatronic designs for a specific instrument from Graston Technique(Model GT3), which is a frequently used tool to clinically deliver localize pressure to the soft tissue. The first design is based on compression load cells, where 4-load cells are used to measure the force components in three-dimensional space. The second design uses a 3D load cell, which can measure all three force components force simultaneously. Both designs are implemented with IMUduino microcontroller chips which can also measure tool orientation angles and provide computed stroke frequency. Both designs, which were created using Creo CAD platform, were also analyzed thorough strength and integrity using the finite element analysis package ANSYS. Once the static analysis was completed, a dynamic model was created for the first design to simulate IASTM practice using the GT-3 tool. The deformation and stress on skin were measured after applying force with the GT-3 tool. Additionally, the relationship between skin stress and the load cell measurements has been investigated. The second design of the mechatronic IASTM tool was validated for force measurements using an electronic plate scale that provided the baseline force values to compare with the applied force values measured by the tool. The load cell measurements and the scale readings were found to be in agreement within the expected degree of accuracy. The stroke frequency was computed using the force data and determining the peaks during force application. The orientation angles were obtained from the built-in sensors in the microchip.Item Injector Waveform Monitoring of a Diesel Engine in Real-Time on a Hardware in the Loop Bench(2011-12) Farooqi, Quazi Mohammed Rushaed; Anwar, Sohel; Wasfy, Tamer; Lee, Jaehwan (John)This thesis presents the development, experimentation and validation of a reliable and robust system to monitor the injector pulse generated by an Engine Control Module (ECM) and send the corresponding fueling quantity to the real-time computer in a closed loop Hardware In the Loop (HIL) bench. The system can be easily calibrated for different engine platforms as well. The fueling quantity that is being injected by the injectors is a crucial variable to run closed loop HIL simulation to carry out the performance testing of engine, aftertreatment and other components of the vehicle. This research utilized Field Programmable Gate Arrays (FPGA) and Direct Memory Access (DMA) transfer capability offered by National Instruments (NI) Compact Reconfigurable Input-Output (cRIO) to achieve high speed data acquisition and delivery. The research was conducted in three stages. The first stage was to develop the HIL bench for the research. The second stage was to determine the performance of the system with different threshold methods and different sampling speeds necessary to satisfy the required accuracy of the fueling quantity being monitored. The third stage was to study the error and its variability involved in the injected fueling quantity from pulse to pulse, from injector to injector, between real injector stators and cheaper inductor load cells emulating the injectors, over different operating conditions with full factorial design of experimentation and mixed model Analysis Of Variance (ANOVA). Different thresholds were experimented to find out the best thresholds, the Start of Injection (SOI) threshold and the End of Injection (EOI) threshold that captured the injector “ontime” with best reliability and accuracy. Experimentation has been carried out at various data acquisition rates to find out the optimum speed of data sampling rate, trading off the accuracy of fueling quantity. The experimentation found out the expected error with a system with cheaper solution as well, so that, if a test application is not sensitive to error in fueling quantity, a cheaper solution with lower sampling rate and inductors as load cells can be used. The statistical analysis was carried out at highest available sampling rate on both injectors and inductors with the best threshold method found in previous studies. The result clearly shows the factors that affect the error and the variability in the standard deviations in error; it also shows the relation with the fixed and random factors. The real-time application developed for the HIL bench is capable of monitoring the injector waveform, using any fueling ontime table corresponding to the platform being tested, and delivering the fueling quantity in real-time. The test bench made for this research is also capable of studying injectors of different types with the automated test sequence, without occupying the resource of fully capable closed loop test benches for testing the ECM unctionality.Item Modeling and control of hydraulic wind power transfer systems(2014) Vaezi, Masoud; Izadian, Afshin; Anwar, Sohel; Zhu, LikunHydraulic wind power transfer systems deliver the captured energy by the blades to the generators differently. In the conventional systems this task is carried out by a gearbox or an intermediate medium. New generation of wind power systems transfer the captured energy by means of high-pressure hydraulic fluids. A hydraulic pump is connected to the blades shaft at a high distance from the ground, in nacelle, to pressurize a hydraulic flow down to ground level equipment through hoses. Multiple wind turbines can also pressurize a flow sending to a single hose toward the generator. The pressurized flow carries a large amount of energy which will be transferred to the mechanical energy by a hydraulic motor. Finally, a generator is connected to the hydraulic motor to generate electrical power. This hydraulic system runs under two main disturbances, wind speed fluctuations and load variations. Intermittent nature of the wind applies a fluctuating torque on the hydraulic pump shaft. Also, variations of the consumed electrical power by the grid cause a considerable load disturbance on the system. This thesis studies the hydraulic wind power transfer systems. To get a better understanding, a mathematical model of the system is developed and studied utilizing the governing equations for every single hydraulic component in the system. The mathematical model embodies nonlinearities which are inherited from the hydraulic components such as check valves, proportional valves, pressure relief valves, etc. An experimental prototype of the hydraulic wind power transfer systems is designed and implemented to study the dynamic behavior and operation of the system. The provided nonlinear mathematical model is then validated by experimental result from the prototype. Moreover, this thesis develops a control system for the hydraulic wind power transfer systems. To maintain a fixed frequency electrical voltage by the system, the generator should remain at a constant rotational speed. The fluctuating wind speed from the upstream, and the load variations from the downstream apply considerable disturbances on the system. A controller is designed and implemented to regulate the flow in the proportional valve and as a consequence the generator maintains its constant speed compensating for load and wind turbine disturbances. The control system is applied to the mathematical model as well as the experimental prototype by utilizing MATLAB/Simulink and dSPACE 1104 fast prototyping hardware and the results are compared.Item Structural Optimization of Thin Walled Tubular Structure for Crashworthiness(2014) Shinde, Satyajeet Suresh; Tovar, Andrés; Anwar, Sohel; Wasfy, TamerCrashworthiness design is gaining more importance in the automotive industry due to high competition and tight safety norms. Further there is a need for light weight structures in the automotive design. Structural optimization in last two decades have been widely explored to improve existing designs or conceive new designs with better crashworthiness and reduced mass. Although many gradient based and heuristic methods for topology and topometry based crashworthiness design are available these days, most of them result in stiff structures that are suitable only for a set of vehicle components in which maximizing the energy absorption or minimizing the intrusion is the main concern. However, there are some other components in a vehicle structure that should have characteristics of both stiffness and flexibility. Moreover, the load paths within the structure and potential buckle modes also play an important role in efficient functioning of such components. For example, the front bumper, side frame rails, steering column, and occupant protection devices like the knee bolster should all exhibit controlled deformation and collapse behavior. This investigation introduces a methodology to design dynamically crushed thin-walled tubular structures for crashworthiness applications. Due to their low cost, high energy absorption efficiency, and capacity to withstand long strokes, thin-walled tubular structures are extensively used in the automotive industry. Tubular structures subjected to impact loading may undergo three modes of deformation: progressive crushing/buckling, dynamic plastic buckling, and global bending or Euler-type buckling. Of these, progressive buckling is the most desirable mode of collapse because it leads to a desirable deformation characteristic, low peak reaction force, and higher energy absorption efficiency. Progressive buckling is generally observed under pure axial loading; however, during an actual crash event, tubular structures are often subjected to oblique impact loads in which Euler-type buckling is the dominating mode of deformation. This undesired behavior severely reduces the energy absorption capability of the tubular structure. The design methodology presented in this paper relies on the ability of a compliant mechanism to transfer displacement and/or force from an input to desired output port locations. The suitable output port locations are utilized to enforce desired buckle zones, mitigating the natural Euler-type buckling effect. The problem addressed in this investigation is to find the thickness distribution of a thin-walled structure and the output port locations that maximizes the energy absorption while maintaining the peak reaction force at a prescribed limit. The underlying design for thickness distribution follows a uniform mutual potential energy density under a dynamic impact event. Nonlinear explicit finite element code LS-DYNA is used to simulate tubular structures under crash loading. Biologically inspired hybrid cellular automaton (HCA) method is used to drive the design process. Results are demonstrated on long straight and S-rail tubes subject to oblique loading, achieving progressive crushing in most cases.Item Towards commercialization of self-healing technology in epoxy coating(2014) Ye, Lujie; Jones, Alan S.; Zhang, Jing; Zhu, Likun; Chen, JieThis work is focused on developing viable self-healing coatings, especially considering the viability of the coating in a commercial context. With this in mind, finding low cost healing agents, with satisfactory healing and mechanical properties as well as adapting the healing system for use in coatings was required. Seven potential healing agents were evaluated and an air-drying triglyceride (linseed oil) was identified as the candidate healing agent. Different encapsulation techniques were evaluated and ureaformaldehyde microcapsules were chosen as the candidate encapsulation technique. Self-healing coatings were fabricated using urea-formaldehyde encapsulated linseed oil. EIS, SEM and TGA technologies were used to evaluate mechanical performance, corrosion resistance, and self-healing performance.Item Video anatomy : spatial-temporal video profile(2014-07-31) Cai, Hongyuan; Zheng, Jiang Yu; Tuceryan, Mihran; Popescu, Voicu Sebastian; Tricoche, Xavier; Prabhakar, Sunil; Gorman, William J.A massive amount of videos are uploaded on video websites, smooth video browsing, editing, retrieval, and summarization are demanded. Most of the videos employ several types of camera operations for expanding field of view, emphasizing events, and expressing cinematic effect. To digest heterogeneous videos in video websites and databases, video clips are profiled to 2D image scroll containing both spatial and temporal information for video preview. The video profile is visually continuous, compact, scalable, and indexing to each frame. This work analyzes the camera kinematics including zoom, translation, and rotation, and categorize camera actions as their combinations. An automatic video summarization framework is proposed and developed. After conventional video clip segmentation and video segmentation for smooth camera operations, the global flow field under all camera actions has been investigated for profiling various types of video. A new algorithm has been designed to extract the major flow direction and convergence factor using condensed images. Then this work proposes a uniform scheme to segment video clips and sections, sample video volume across the major flow, compute flow convergence factor, in order to obtain an intrinsic scene space less influenced by the camera ego-motion. The motion blur technique has also been used to render dynamic targets in the profile. The resulting profile of video can be displayed in a video track to guide the access to video frames, help video editing, and facilitate the applications such as surveillance, visual archiving of environment, video retrieval, and online video preview.