Rapid Prototyping System Research Articles

End-to-End Deep Neural Network Based Nonlinear Model Predictive Control: Experimental Implementation on Diesel Engine Emission Control

In this paper, a deep neural network (DNN)-based nonlinear model predictive controller (NMPC) is demonstrated using real-time experimental implementation. First, the emissions and performance of a 4.5-liter 4-cylinder Cummins diesel engine are modeled using a DNN model with seven hidden layers and 24,148 learnable parameters created by stacking six Fully Connected layers with one long-short term memory (LSTM) layer. This model is then implemented as the plant model in an NMPC. For real-time implementation of the LSTM-NMPC, an open-source package acados with the quadratic programming solver HPIPM (High-Performance Interior-Point Method) is employed. This helps LSTM-NMPC run in real time with an average turnaround time of 62.3 milliseconds. For real-time controller prototyping, a dSPACE MicroAutoBox II rapid prototyping system is used. A Field-Programmable Gate Array is employed to calculate the in-cylinder pressure-based combustion metrics online in real time. The developed controller was tested for both step and smooth load reference changes, which showed accurate tracking performance while enforcing all input and output constraints. To assess the robustness of the controller to data outside the training region, the engine speed is varied from 1200 rpm to 1800 rpm. The experimental results illustrate accurate tracking and disturbance rejection for the out-of-training data region. At 5 bar indicated mean effective pressure and a speed of 1200 rpm, the comparison between the Cummins production controller and the proposed LSTM-NMPC showed a 7.9% fuel consumption reduction, while also decreasing both nitrogen oxides (NOx) and Particle Matter (PM) by up to 18.9% and 40.8%.

Development and verification of model predictive control strategy for selective catalytic reduction with rapid prototyping

The International Maritime Organization has set forth strict regulations about the NOx emission limit of ships to address the pollution problem in the transportation sector. Selective catalytic reduction (SCR) is the most effective post-engine means to reduce emission, and the control of urea spray critically affects the SCR performance. This paper developed a predictive control strategy of the SCR system based on the state space model. Specifically, the block model of the SCR reactor was established based on the reaction mechanism of the SCR system, and the unknown model parameters were identified. Rapid prototyping of the SCR system was carried out using the dSPACE real-time simulation platform to optimize the control under steady-state conditions. The catalyst ammonia coverage was set as the control target to maximize the NOx conversion rate while maintaining the NH3 escape at the outlet less than 10 ppm. The cyclic test results showed that the control strategy can quickly respond to external disturbance and modify the control target to keep the ammonia coverage rate within the target range, to thus achieve high NOx conversion rate while maintaining the required ammonia escape.

Design and development of a rapid prototyping system combining traditional fused deposition modeling and reconfigurable pins platform

Additive manufacturing, also commonly referred to as 3D printing, has enabled the rapid production of parts possessing complex geometry. Despite Fused Deposition Modeling (FDM) being the most popular process, it still suffers from certain limitations, such as production time and material required to build the material support structure for overhang parts. This results in increased total build time and material requirement. This article aims to eliminate these limitations by using a proposed hybrid reconfigurable pin FDM system. A hybrid system is proposed to eliminate the problem of support build time and material requirements. The proposed hybrid FDM is the combination of traditional FDM with a reconfigurable pins platform. The proposed hybrid system has been implemented for printing an overhang test part. The result analysis also reveals that the proposed hybrid system is an effective way of producing overhang parts without material support structures. A real-life case example is considered for validation of the proposed reconfigurable pins platform.

Parameterisation of domain knowledge for rapid and iterative prototyping of knowledge-based systems

In critical infrastructure applications, timely and consistent fault detection and diagnosis is an increasingly important operational process, especially in the energy sector where safety is of the utmost importance. To realise this, engineers have to manually analyse data acquired from several assets using predefined diagnostic processes, but this is a time-consuming process requiring significant amounts of specialist expert knowledge. Data-driven approaches to support fault detection and diagnosis, and other similar problems, can produce accurate results comparable to what the engineers can achieve in a fraction of the time. However, the majority of these data-driven techniques are black box techniques and lack explainability which is often necessary for explaining decisions about critical assets in the power generation industry. Knowledge-based systems, such as rule-based expert systems have been shown to provide not only accurate decisions but also the explanation and reasoning behind these decisions in some related applications. However, there is a significant time cost associated with the development of knowledge-based systems, and in particular with the knowledge elicitation process, where the domain expert’s knowledge is formalised and is encoded into the system. This challenge is commonly referred to as the knowledge elicitation bottleneck.In this paper, we present a novel approach to performing the knowledge elicitation using a set of symbolic primitives (rise, fall, fluctuate, and stable) to parameterise typical time-series condition monitoring data. The knowledge is represented by using a common language that can easily be communicated with (and from) the domain experts. This allows for the quick and accurate elicitation of the domain experts knowledge, but also the formalisation and implementation of the knowledge into a rapidly produced diagnostic system. Further to this, due to the parametrisation of the knowledge, it is possible to iteratively improve the knowledge base by updating these parameters based on new unseen data. This approach was applied to the Tennessee Eastman dataset, which is simulated data of a real-world industrial process. It was found that by using this approach it was possible to accurately and quickly capture the knowledge required to detect several faults within the case study dataset, but also provided fully explained reasons why each fault was detected by relating the explanations to the symbolic primitives previously defined.

Combustion rate shaping for flex-fuel applications

Compliance with future greenhouse gas (GHG) and pollutant emissions poses major challenges for the further development of advanced internal combustion engines for light and heavy-duty applications. To meet all standards, the use of synthetic fuels and/or E-Fuels is an important alternative due to their enormous potential in terms of energy density, sustainability and low pollution combustion. However, the increased fuel diversity associated with the use of these alternative fuels adds an additional layer of complexity to the powertrain development and calibration process, resulting in an increase in development time and cost. This necessitates the application of advanced model-based closed-loop control strategies to optimize the fuel- and air-path calibration to make best use of the properties of the different fuels. One potential solution to address the impact of flex-fuel variances on the combustion process is the Combustion Rate Shaping (CRS) concept presented in this paper. It can maintain a desired combustion trace irrespective of fuel variations, hardware drifts and other external factors, thus ensuring optimal combustion. The highest benefit of this control concept comes from its application directly on the engine control unit (ECU) in combination with an in-cylinder measurement. This enables the online real-time control of the combustion, regardless of which fuel is being used. In addition, this also enables optimization and adaption of the fuel- and air-path settings to the used fuel while driving the vehicle. However, so far, the high computational cost of this control concept limits its real-time capability. Therefore, the optimization of the control concept to achieve real-time capability is the focus of this paper. The optimized control strategy is implemented on a demonstrator engine test bench built using a Rapid Control Prototyping (RCP) system. The controller performance is demonstrated both under steady-state and transient operating conditions, and the potential of the concept is demonstrated in several cases such as engine temperature variations, injector drift, individual cylinder balancing, and fuel-variation.

Low-fidelity design optimization and development of a VTOL swarm UAV with an open-source framework

Hybrid configurations in aircraft design are highly favorable as they can achieve the appropriate trade-offs required to develop a generalized unmanned aerial system (UAS). Rapid prototyping of such systems at the student level is challenging because commercial software is expensive and difficult to interlink with other tools for creating a multi-disciplinary design. We address this challenge by conceptualizing an aircraft design framework made entirely of open-source software, libraries, and in-house code. We then use this framework to design an all-electric unmanned aerial system with transitioning Vertical Take-off and Landing (VTOL) and Fixed-Wing (FW) modes. The UAV is capable of long-range surveillance up to 100 Kilometers and carrying a maximum relief payload of 1 kg while operating in an ad-hoc wi-fi network with a swarm of similar UAVs. A low fidelity particle swarm optimization algorithm (PSO) and a comprehensive propulsion architecture is also incorporated and validated against commercial software. To validate the design, a prototype is fabricated from glass-fiber and XPS foam, integrated with appropriate sensors and tuned using ArduPilot software. The results show that low-fidelity design is a safe starting point for prototyping under constrained timelines. The study is concluded by discussing the technical challenges of using free software, and some practical considerations while flight testing a UAV with a hybrid configuration.

Development of a digital twin for real-time simulation of a combustion engine-based power plant with battery storage and grid coupling

Coordinated control of combustion engine-based power plants with battery storage is the next big thing for optimising renewable energy. Digital twins can enable such sophisticated control but currently are too simplistic for the required insight. This study explores the feasibility of a fully physics-based combustion engine model in real-time co-simulation with an electrical power plant model, including battery storage. A detailed, crank-angle resolved, one-dimensional model of a large-bore stationary engine is reduced to a fast-running model (FRM). This engine digital twin is coupled with a complete power plant control model, developed in Simulink. Real-time functions are tested on a dedicated rapid-prototyping system using a target computer. Measurement data from the corresponding power plant infrastructure provide validation for the digital twin. The model-in-the-loop simulations show real-time results from both the standalone combustion and electric submodels mostly within 5% of measured values. The model coupling for fully predictive simulation was tested on a desktop computer, showing expected functionality and validity within 4% and 8% of the respective measured generator and converter outputs. However, execution time of the FRM needs reducing when moving to final hardware-in-the-loop implementation of a complete power plant model.

Optimization of plotter printing for sub-terahertz metallic metasurfaces fabrication on ultra-thin substrate

Additive manufacturing processes have emerged as a promising way to conceive terahertz and millimetric components. In this work, we discuss a printing process for sub-terahertz metallic metasurfaces fabrication on ultra-thin substrates. We demonstrate the use of a plastic substrate with a micrometric thickness which makes this printing method a promising and low cost alternative to conventional optical lithography for the fabrication of flexible terahertz 2D metasurfaces. After detailing the key parameters and limitations, we applied the optimized process to fabricate samples composed of periodic arrays of split ring resonators on 50 µm thick fused silica and 3 µm thick PET substrates. The optical response in transmission of the metasurfaces shows expected resonances in the 100 GHz range and demonstrates the use of microplotter system for rapid prototyping of low-loss terahertz passive components on ultra-thin substrate.

Rapid DNA origami nanostructure detection and classification using the YOLOv5 deep convolutional neural network

The intra-image identification of DNA structures is essential to rapid prototyping and quality control of self-assembled DNA origami scaffold systems. We postulate that the YOLO modern object detection platform commonly used for facial recognition can be applied to rapidly scour atomic force microscope (AFM) images for identifying correctly formed DNA nanostructures with high fidelity. To make this approach widely available, we use open-source software and provide a straightforward procedure for designing a tailored, intelligent identification platform which can easily be repurposed to fit arbitrary structural geometries beyond AFM images of DNA structures. Here, we describe methods to acquire and generate the necessary components to create this robust system. Beginning with DNA structure design, we detail AFM imaging, data point annotation, data augmentation, model training, and inference. To demonstrate the adaptability of this system, we assembled two distinct DNA origami architectures (triangles and breadboards) for detection in raw AFM images. Using the images acquired of each structure, we trained two separate single class object identification models unique to each architecture. By applying these models in sequence, we correctly identified 3470 structures from a total population of 3617 using images that sometimes included a third DNA origami structure as well as other impurities. Analysis was completed in under 20 s with results yielding an F1 score of 0.96 using our approach.

Robust control technique in power converter with linear induction motor

Induction motors are widely used in industrial applications as actuators, thanks to their simplicity of construction, which is subsequently reflected in low-cost maintenance. This paper shows the behavior of a three-phase power converter with a linear induction motor (LIM) as load, using a quasi-sliding control technique for output voltage regulation and a new control technique to control chaos. Digital pulse width modulation (DPWM) techniques are widely used to control electronic power converters. The controller proposed in this paper was designed using zero average dynamic (ZAD) and fixed point inducting control (FPIC) techniques. The ZAD-FPIC control strategy was designed and applied to a three-phase converter with linear induction motor load. Since it is not possible to measure the secondary currents, a secondary current observer was included in the system. Finally, bifurcation diagrams are shown as a technique for tuning controller parameters in ZAD-FPIC controllers. For the illustration of numerical results a simulation of the linear induction motor drive controlled was made by MATLAB/Simulink. The designs were tested in a rapid control prototyping (RCP) system based on digital signal processing (DSP) for dSPACE platform, using the 1103 controller card and control desk interface.

Optimization Techniques for Online MPC in Android Smartphones for Artificial Pancreas: A Comparison Study

Model Predictive Control (MPC) has evolved as a potential approach for safety-critical Artificial Pancreas (AP) systems because of its constraint handling capability. However, it demands a solution of a fairly high-dimensional constrained optimization problem online. A popular approach to close the loop in practice is to implement it in smartphones in general and in Android smartphones in particular (because of its pervasiveness). As MPC requires the use of Optimization solvers in smartphones, a survey of various available options has been done in this paper, followed by a thorough comparison study about their performance. This comprehensive comparison includes the ease of convergence of the optimization solution and the computational time. It turns out that the Python-based optimization solver SciPy is most suitable for this purpose as compared to other alternatives such as Casadi, lpsolvers and autocode generation process of MATLAB. The outcome of this study can be used for rapid prototyping of Android smartphone-based AP systems.

Mixed Reality in the Reconstruction of Orbital Floor: An Experimental and Clinical Evaluative Study.

Introduction:Orbital floor fractures are common within midface fractures. Their management includes restoration of orbital volume and anatomy. However, these procedures could be associated with the mispositioning of implants and inadequate volume restoration. Nowadays medical rapid prototyping, virtual planning (VP), and navigation systems significantly increase the precision of such procedures. Nevertheless, the application of intraoperative navigation could be associated with intraoperative mistakes related to two-dimensional imaging. The application of mixed reality (MR) could solve this problem. The current study aims to demonstrate the application of MR in orbital reconstruction.Materials and Methods:The current study included experimental and clinical implementation of MR in orbital reconstruction. Within the experimental part, 10 residents and 5 experienced maxillofacial surgeons were added. All data and customised software were well documented and then used in a single clinical case of orbital floor reconstruction.Results:Visual assessment of plate positioning within the experiment revealed proper plate positioning in 8 cases. A comparison of virtual and real measurements showed a stable deviation of 0.65–1.15 (mean 0.9 mm). As a result of the clinical implementation of MR technology, after surgical reconstruction, the patient showed improvement in ocular mobility and reduction of diplopia. A postoperative computed tomography scan showed proper plate positioning.Discussion:Implementation of MR based on VP could significantly improve the results of preoperative planning, intraoperative navigation, and surgery. However, existing technical limitations that relate to navigation principles could produce mistakes and errors. Therefore, further investigations related to the 6 degrees of freedom problem solution are considered reasonable in the elimination of listed issues.

A Unified Active Damping for Grid and Converter Current Feedback in Active Front End Converters

The Active Front End (AFE) is interconnected to the grid with an LCL filter to reduce the switching harmonics. Possible resonances induced by the LCL filter can be damped using Active Damping (AD). In AFE applications, the current sensor can be placed on either the grid-side or the converter-side, resulting in different frequency responses. This paper presents a novel active damping approach that is universally applicable to all active front end converters regardless of whether the current sensor is placed on the grid side or the converter side. At the heart of the proposed approach is a practically implementable fourth-order filter that dampens the resonance whilst being capable of mitigating the challenges associated with both grid current feedback and converter current feedback. These include amplifications associated with derivatives in active damping term with grid current feedback and high susceptibility to instability with converter current feedback due to digital delay. The proposed active damping is thoroughly investigated in terms of practical implementation issues such as different LCL filter parameters, different sampling methods, grid impedance, and digital delay. The approach is experimentally validated on motor drives consisting of 2 two-level three-phase converters interfaced with a dSPACE rapid control prototyping system across a wide range of resonant frequencies and PWM sampling methods. Stability analysis is performed on both current control methods to validate that active damping assures stable operation for a wide range of resonant frequencies and grid impedance.

AI-Aided Online Adaptive OFDM Receiver: Design and Experimental Results

Orthogonal frequency division multiplexing (OFDM) has been widely applied in many wireless communi- cation systems. The artificial intelligence (AI)-aided OFDM receivers are currently brought to the forefront to replace and improve the traditional OFDM receivers. In this paper, we first compare two AI-aided OFDM receivers, namely, data-driven fully connected deep neural network and model-driven ComNet, through extensive simulation and real-time video transmission using a 5G rapid prototyping system for an over-the-air (OTA) test. We find a performance gap between the simulation and the OTA test caused by the discrepancy between the channel model for offline training and the real environment. We develop a novel online training system, which is called SwitchNet receiver, to address this issue. This receiver has a flexible and extendable architecture and can adapt to real channels by training only several parameters online. From the OTA test, the AI-aided OFDM receivers, especially the SwitchNet receiver, are robust to OTA environments and promising for future communication systems. At the end of this paper, we discuss potential challenges and future research inspired by our initial study in this paper.

Learning and Playing in Early Childhood with Augmented Reality Technology

Purpose: Helping the learning process in early childhood through playing and learning activities with Augmented Reality technology.Design/methodology/approach: Using Augmented Reality technology with the Iterative Rapid Paper Prototype system development methodFindings/result: Based on tests conducted on 5 types of android devices, 10 samples of early childhood participants (4-5 years) and 5 groups of objects consisting of 10 types resulted in an increase in learning ability of 33.35% which was sourced from the measurement of the correct answers that were successfully obtained. between learning methods through pictures and learning using Augmented Reality technologyOriginality/value/state of the art: In previous research, the learning model was carried out on elementary school children (aged 6 years and over) and without the implementation of Augmented Reality technology

A critical study and analysis of process parameters of selective laser sintering Rapid prototyping

In the Rapid Prototyping process, surface finish is very critical as it affects the part quality. As Rapid Prototyping is moving towards Rapid manufacturing there is an increasing stress on obtaining good quality parts. Surface finish of Rapid Prototyping parts is becoming more and more important with more parts being used for end purposes. Good surface finish is required for better functionality and look, but also for cost reduction in terms of reducing post-processing expenditure. A critical study and analysis of process parameters behaviour has been carried out for assessing the surface quality of parts manufactured on Selective Laser Sintering Rapid Prototyping system. The principal concern of experimental work has been investigated for achieving better surface finish by which the life of parts is governed. The primary objective of study is to analyze the optimum process parameters of Selective Laser Sintering Rapid Prototyping system. The material selected for study is “Dura Form“ and experimental work was carried out on selective laser sintering rapid prototyping machine DTM Vanguard 2500. Taguchi method is used to investigate the parameters and results are improved by using signal-to-noise methods. The major process parameters selected for study were Part Orientation, Laser Power and Scan Spacing. The study and analysis depict that Part Orientation has a major effect on the surface roughness of the part produced on Rapid Prototyping Machine. It has been observed from the study that lower value of Part Orientation level I (0 degree) gives better finish and this is because of less number of layers required for manufacturing of part. It has been observed from the study of second process parameter i.e. Laser Power that the surface finish was better at 16 W. Similarly, the surface finish obtained at lower values of Scan Spacing was for better as compared to higher values of Scan Spacing. The comparative analysis of three-process parameter indicates that the effects of variation of Part Orientation and Scan Spacing parameters are more as compare to Laser Power.

Real-Time Integrated Power and Thermal Management of Connected HEVs Based on Hierarchical Model Predictive Control

Cabin heating requirement and engine efficiency degradation in cold weather lead to considerable increase in fuel consumption of hybrid electric vehicles (HEVs). This article focuses on the development of a real-time control framework for integrated power and thermal management (i-PTM) of a connected HEV through incorporating vehicle speed preview informed by traffic connectivity. To evaluate the ceiling of the fuel economy improvement via i-PTM, forward dynamic programming (FDP) is adopted as a benchmark study to find the global optimization solution and then implemented into a centralized model predictive controller (CMPC). Regarding the slow response associated with the coupled electric-thermal dynamics in the HEV, a two-layer hierarchical MPC (HMPC) framework is developed, which exploits vehicle speed preview predictions over short and long prediction horizons to optimize fuel consumption while satisfying power and cabin heating demand. The simulation results show that the proposed HMPC provides near-optimal solutions in reference to the benchmark while reduces up to 79% computation burden compared with CMPC. Additionally, compared to a baseline power management (PM) controller, up to 5.46% improvement on fuel economy can be achieved by HMPC for congested driving scenarios. Finally, the real-time feasibility of the HMPC is assessed in a dSPACE rapid prototyping system.

Fabrication of medium scale 3D components using a stereolithography system for rapid prototyping

A cost-effective stereolithography for medium-scale components is developed to fabricate 3D components with high build speed and resolution from photo-curable resin. The developed SLA utilizes a focused light beam of wavelength range (300 nm – 700 nm) coming from the DLP projector, passes through the objective lens, and finally is imposed on the platform containing a photo-curable resin layer. After focusing the light beam on the liquid resin layer, the photo-polymerization reaction occurs, and the liquid resin becomes solid. Thus, the 3D object is fabricated layer by layer, curing of liquid resin. The photopolymer used in this experiment is polyethylene glycol di-acrylate, and Irgacure 784 as photo-initiator. The Creo 3.0 software is used for the modeling of 3D objects. A special MATLAB code is developed for slicing of the 3D model and displaying the sliced image one by one through the DLP projector. The Arduino microcontroller with a stepper motor and ball screw is used to control the motion of the Z-stage platform. The Creation workshop software is also used to control the motion of the Z-stage and period to display the sliced images through the DLP projector. The medium-scale 3D objects with rectangular, square, and circular cross-sections are obtained by curing the aforementioned photo-curable resin. It is observed that the 3D objects are best cured for two seconds curing time with 0.1 mm curing depth along Z-axis.

Facing COVID-19 pandemic: development of custom-made face mask with rapid prototyping system.

COVID-19 is a global pandemic. The virus spreads through respiratory droplets and close contact. Therefore, the availability of personal protective equipment (PPE) for healthcare professionals is essential. 3D printing technology could represent a valid option to ameliorate PPE shortages. Custom-made face mask were designed on the basis of facial scan and then 3D-printed. The whole protocol is executed with freeware software and only required a 3D printer. Six healthcare workers wore the device weekly thus expressing a judgment regarding quality of work, respiratory and skin comfort. The estimated total cost of a single mask is approximately 5 USD. The virtual design of a complete mask lasted 68 minutes on average. Most healthcare workers rated comfort as very good. Based on the encouraging results obtained, we can confidently confirm that custom-made masks are novel and useful devices that may be used in the fight against COVID-19.

Functionalized Magnetic 3D Printing of High Performance Electrodes for Lithium Battery

With the trend for rapid prototyping of parts and systems, additive manufacturing process like three-dimensional (3D) printing can advance the fabrication of shape-specific architectures for anisotropic permanent magnets involving expensive and critical rare-earth elements1,2. Although in-situ alignment will open new horizons for manufacturing of complex permanent magnets, integration of alignment systems into 3D printing processes is still a challenge. In our former works, we have demonstrated a mechanism to address this challenge accomplished by integration of bonded magnet production, 3D printing, finite element modeling and system development3.Alignment of electrode pores, (e.g. using magnetic field) allow from better electrochemical transport of electrolyte through the electrode, thereby improving cell performance. The formerly developed magnetic field alignment model for 3D printed material in my postdoctoral program will be modified and used to predict the degree-of-alignment in electrodes for lithium-ion battery. The relationship between pore tortuosity and degree-of-alignment was modeled using a power law with the effect of exponent compared in Fig. 1a.A coupled electrochemical-thermal-mechanical modeled is developed to predict the cell capacity loss due to above-mentioned anodic degradation mechanisms. Keeping the porosity, the tortuosity is altered between aligned and non-aligned states, to evaluate the electrochemical transport properties like diffusivity and ionic conductivity in the electrode. The model predicts an increase in cell capacity for lower tortuosity (aligned) electrodes, which is further amplified at faster charging conditions (Fig. 1b). It is generally observed that fast charging induces a steep non-uniform charge distribution in the unaligned electrodes. With increase in alignment, the electrolyte has easier pathways leading to a more uniform charge distribution and therefore a larger capacity. An immediate impact of a more uniform charge distribution is the lowering of tendency for lithium-ions to deposit in metallic form on the anode (called lithium plating).These results provide an indication that alignment of electrodes will be beneficial towards designing a higher capacity and faster charging battery. Experimental validations are the next step towards validating the model predictions. Huber, C. et al. 3D print of polymer bonded rare-earth magnets, and 3D magnetic field scanning with an end-user 3D printer. Appl. Phys. Lett. 109, 162401 (2016).Li, L., Post, B., Kunc, V., Elliott, A. M. & Paranthaman, M. P. Additive manufacturing of near-net-shape bonded magnets: Prospects and challenges. Scr. Mater. (2017) doi:10.1016/j.scriptamat.2016.12.035.Sarkar, A. et al. Functionalizing magnet additive manufacturing with in-situ magnetic field source. Addit. Manuf. 34, 101289 (2020). Figure 1