Jagdish Chandra Patra

Cold spray additive manufacturing (CSAM) has recently gained increased attention from the research community and industries due to its several advantages, such as high metal deposition rate, low working temperature, and ability to deposit high-reflectivity materials. Despite these advantages, one of the main limitations of CSAM is poor dimensional accuracy of as-fabricated components. Hence, these components often require costly time-consuming postfabrication machining. Poor dimensional accuracy is mainly attributed to the lack of proper control of spray deposit profiles (SDPs). The shape of the SDP is mainly controlled by a set of input process parameters (IPPs), e.g., spray angle, nozzle standoff distance, and velocity. However, there is no systematic procedure or technique available to determine the IPPs a priori to achieve a desired SDP. In this article, we propose a machine learning-based hybrid multilayer perceptron and particle swarm optimization inverse modeling technique to predict a set of IPPs to achieve a desired SDP. Using a dataset obtained from the literature, with extensive simulation studies, we have provided evidence of accuracy, stability, and reliability of prediction of IPPs in terms of prediction error, convergence characteristics, and consistent results. Considering highly nonlinear complex mapping between the IPPs and the SDP, we believe the performance of the inverse model to be satisfactory.

Smart city management is facing a new challenge from littered face masks during COVID-19 pandemic. Addressing the issues of detection and collection of this hazardous waste that is littered in public spaces and outside the controlled environments, usually associated with biomedical waste, is urgent for the safety of the communities around the world. Manual management of this waste is beyond the capabilities of governments worldwide as the geospatial scale of littering is very high and also because this contaminated litter is a health and safety issue for the waste collectors. In this paper, an autonomous biomedical waste management framework that uses edge surveillance and location intelligence for detection of the littered face masks and predictive modelling for emergency response to this problem is proposed. In this research a novel dataset of littered face masks in various conditions and environments is collected. Then, a new deep neural network architecture for rapid detection of discarded face masks on the video surveillance edge nodes is proposed. Furthermore, a location intelligence model for prediction of the areas with higher probability of hazardous litter in the smart city is presented. Experimental results show that the accuracy of the proposed model for detection of littered face masks in various environments is 96%, while the speed of processing is ten times faster than comparable models. The proposed framework can help authorities to plan for timely emergency response to scattering of hazardous material in residential environments.

Recently the circuit under pad (CUP) designs are getting very popular because they could save lots of silicon especially for high density and high complexity applications. Polysilicon is a material widely used in many electronic device components. When such devices are placed underneath the bond pads, it is highly expected not to show any change in the electrical behavior due to any kind of thermo-mechanical stresses such as probing or wire bonding mainly during the process of packaging. This paper presents a detailed investigation of such impacts on a polysilicon resistor placed underneath the bond pads. The standard meander-shaped polysilicon resistor was used as a test vehicle. In this paper, different stress-causing factors like probing insertions, probing temperature, and wire bond materials were investigated in detail. The impact of such thermo-mechanical stresses is simulated using Comsol to understand the stress phenomenon and for qualitative analysis. Then, the wafer-level silicon testing was carried out to quantify the impact of such thermo-mechanical stresses using a standard Design of Experiments (DOE). The impact in terms of change in resistance of the polysilicon resistor of two different widths (5 μm and 0.42 μm) for two different bond pad metal options viz METMID (thickness of 880 nm) and METTHK (thickness of 3000 nm) was investigated. The percent change in resistance was observed very negligible after multiple probing attempts and probe temperatures. However, after wire bonding, up to ~14% change in resistance was observed when copper was used as the wire bonding material. Nevertheless, this percent change was still well within the reliability specification limit set by JEDEC, which is widely used in the industry. This detailed investigation will help CUP designers to choose the correct Back End Of Line (BEOL) metal options and also wire bonding material upfront and accordingly plan their design margins.

In this paper a new real-time model order reduction technique for stability prediction in the smart grid is proposed. The proposed method uses an online proper orthogonal decomposition algorithm. A snapshot matrix on a sliding sampling window is used for extracting the main components of the system states by performing a randomized singular value decomposition. After reducing the order of the system, a local linear model is estimated for this snapshot matrix. Then, the state of the system is predicted in a sliding prediction window. Finally, a suitable stability index is calculated and the stability of the system is forecasted in this prediction window. The proposed method is capable of predicting the transient stability, unstable/critical machines and the stability limit. In addition, it can be used for first swing and multi-swing instability detection. The simulations on three test systems show that the proposed technique can predict system stability with high precision in real-time. The computational burden and length of prediction horizon is suitable for practical applications and the proposed algorithm has significant advantages in case of large-scale power systems.

We propose a novel approach called, an orthogonal particle swarm optimization (OPSO) algorithm, for economic dispatch (ED) of thermal generating units (TGUs) in smart electric power gird (SEPG) environment. The characteristics of TGUs are nonlinear and the generation system becomes more and more complicated when these TGUs are subjected to ramp rate constraints and prohibited operating zones. In such case, the cost functions become non-smooth and non-convex due to the discontinuities in the cost curves. Moreover, for large-scale TGUs, the high dimensions used in ED problem become a big challenge to find global minimum and to avoid falling into local minima. The proposed OPSO algorithm has the ability to solve such complex problems including ED. The OPSO algorithm applies an orthogonal diagonalization process. It makes d particles (out of total m particles, m ≥ d) that have the possible solutions by constructing orthogonal vectors in the d-dimensional search space. These orthogonal vectors are generated and updated in each iteration and are utilized to guide the d particles to fly in one direction toward global minimum. The OPSO algorithm is evaluated and tested through 40 TGUs and its performance is compared with several other optimization methods. We found that the OPSO algorithm provides better results in term of cost under power grid constraints. Furthermore, we have shown that the OPSO algorithm significantly improves the PSO algorithm in terms of high solution quality, robustness and convergence.

In recent years, multi‐crystalline solar grade silicon (mc‐SoG‐Si), instead of expensive electronic‐grade Si, is being considered in photovoltaic industry for production of solar modules. These materials usually contain a comparable amount of acceptors (e.g. boron) and donors (e.g. phosphorus) and are therefore called compensated mc‐SoG‐Si. The electrical parameters, e.g. majority carrier mobility ( μ ), majority carrier density ( p ) and resistivity ( ρ ), of compensated mc‐SoG‐Si which affect performance of the solar cells vary non‐linearly with temperature due to several complex mechanisms. In this study, the authors propose artificial neural network (ANN)‐based models to predict the three electrical parameters of mc‐SoG‐Si material. Using a limited amount of measurement data, the authors have shown that the ANN‐based models can predict the three electrical parameters of a given sample over a wide temperature range of 70–400 K and a specific range of compensation ratio. The authors have shown with extensive simulated results that these models can predict the three parameters with a maximum error of ±10%.

We propose an enhanced Particle Swarm Optimization (PSO) algorithm named Orthogonal PSO (OPSO) algorithm, for Economic Dispatch (ED) of the generated power in a smart grid environment. The equality and inequality constraints, and power balance response against mismatch between load demand and total power outputs of generating units involve nonlinear characteristics and non-smooth cost functions. The proposed OPSO algorithm has the ability to solve such complex problems of power systems including ED. The OPSO algorithm applies Orthogonal Vectors (OVs) in the d-dimensional search space. The d particles that have possible solutions move in the d-dimensional search space to form OVs. These OVs are generated and updated in each iteration and they used to guide those particles to fly in one direction toward global minimum. The OPSO algorithm is evaluated and tested through 15 generating units and its performance is compared with several other optimization methods. We found that OPSO algorithm provides better results in solving the total cost, and power constraints. Furthermore, the OPSO algorithm is succeeded to improve the PSO algorithm in terms of high solution quality, robustness and convergence.

Because of economic and energy-consumption considerations, multicrystalline solar grade silicon (mc-SoG-Si), instead of expensive electronic-grade Si, is being considered in photovoltaic (PV) industry for production of solar modules. These materials usually contain a comparable amount of acceptors (e.g., Boron) and donors (e.g., Phosphorus) and are therefore called compensated mc-SoG-Si. The three main electronic parameters, e.g., majority carrier mobility (μ), majority carrier density (p) and resistivity (ρ), of compensated mc-SoG-Si vary nonlinearly with temperature due to several complex mechanisms. In this paper, we propose two artificial neural network (ANN)-based models to predict these electronic parameters of mc-SoG-Si material. Using a limited amount of measurement data, we have shown that the first ANN-based model can predict the three electronic parameters of a given sample without accounting for the compensation ratio over a wide temperature range of 70-400 K. Whereas, the second ANN model can predict these electronic parameters of a given sample with unknown compensation ratio over the same temperature range. With extensive simulation results we have shown that these models can predict the three parameters with a maximum error of ±10%

Recently Particle Swarm Optimization (PSO) algorithms have been used to solve several engineering optimization problems. In this paper, we propose a novel Orthogonal PSO Learning Algorithm (OPSOLA) to solve Economic Dispatch (ED) problems in a Smart Electric Power Grid (SEPG) application. Thermal power “turbine-generator” has several nonlinear characteristics, e.g., the ramp-rate limits, prohibited operating zones and non-smooth cost functions. The proposed OPSOLA has ability to solve such complex problems in SEPG. The OPSOLA utilizes a combined method by adding orthogonality to PSO algorithm. In this combined learning strategy, the particles move and construct a new exemplar to guide the particles to fly more steadily toward the optimum solution. This is accomplished by determining the promising movements of the candidate particle in subsequent iterations based on orthogonality. The OPSOLA is evaluated and tested through a IEEE 6-unit thermal power plant and by comparing the results with several other optimization methods. We found that OPSOLA provides better performance in solving the ED problems in terms of convergence characteristics, quality of solution and execution time.