• Issue front cover thumbnail

      Volume 44, Issue 3

      March 2019

    • ALDL: a novel method for label distribution learning


      More Details Abstract Fulltext PDF

      Data complexity has increased manifold in the age of data-driven societies. The data has become huge and inherently complex. The single-label classification algorithms that were discrete in their operation are losing prominence since the nature of data is not monolithic anymore. There are now cases in machine learningwhere data may belong to more than one class or multiple classes. This nature of data has created the need for new algorithms or methods that are multi-label in nature. Label distribution learning (LDL) is a new way to view multi-labelled algorithms. It tries to quantify the degree to which a label defines an instance. Therefore, for every instance there is a label distribution. In this paper, we introduce a new learning method, namely, angular label distribution learning (ALDL). It is based on the angular distribution function, which is derived from thecomputation of the length of the arc connecting two points in a circle. Comparative performance evaluation in terms of mean-square error (MSE) of the proposed ALDL has been made with algorithm adaptation of k-NN (AA-kNN), multilayer perceptron, Levenberg–Marquardt neural network and layer-recurrent neural network LDL datasets. MSE is observed to decrease for the proposed ALDL. ALDL is also highly statistically significant for the real world datasets when compared with the standard algorithms for LDL.

    • Efficient band selection for improving the robustness of the EMD-based cepstral features


      More Details Abstract Fulltext PDF

      Mel-Frequency Cepstral Coefficients (MFCC) are features widely and successfully used for various speech processing applications. These features are extracted using Fourier transform. However, this transform suffers from some crucial restrictions when used for analyzing nonlinear and non-stationary signals such asspeech. To address this problem, in the present study, we investigate the application of Empirical Mode Decomposition (EMD) in extracting more efficient and robust features for automatic gender identification. In particular, in the proposed approach, the speech signal is first decomposed into a set of narrow-band oscillatory modes, using EMD, from which mel-frequency cepstral features can be extracted. On the other hand, multi-band decomposition of all modes results in some redundant and even irrelevant features that can degrade the performance of the classification. Therefore, we propose to efficiently select the most discriminative frequency bands over all modes. The minimal-redundancy-maximal-relevance (mRMR) feature selection algorithm is also examined for this purpose. The proposed EMD-based features are then extracted by applying DCT on log power values calculated over the selected mel-scale bands of the IMFs. Simulation results show that, using the proposed features for automatic gender identification considerably improves the performance of the system, inparticular in noisy environments.

    • A Fuzzy Decision-Making Trial and Evaluation Laboratory approach to analyse risk factors related to environmental health and safety aspects in the healthcare industry


      More Details Abstract Fulltext PDF

      In healthcare firms, environmental health and safety (EHS) remains as a vital factor as healthcare products pose very intricate problems related to environment safety. The different similar and dissimilar risk factors that prevail in the system have complicated known and unknown causal relationships that are difficult tounderstand and interpret. Hence, improving the EHS remains as a challenge in healthcare industry. A research study is carried out utilizing the data (in conjunction with expert’s opinion) and conditions of a healthcare firm in India to categorize and obtain the prominent risk factors based on identifying the most adverse causal relationship among them. A Fuzzy Decision-Making Trial and Evaluation Laboratory (fuzzy DEMATEL)-based approach is designed and employed to assess and rank different EHS risk factors. The trapezoidal fuzzy membership function of the model facilitates better learning of interrelationships in spite of the prevailing vagueness in the causal relationships between the risk factors. The outcomes (the decisive risk factors) out of the experimentation using the proposed methodology strongly coincide with the actual causes of the EHS factorsduring the last one decade. As the proposed approach is found to be very effective in fixing the causal relationships and ranking among the risk factors, this may be successfully employed in similar healthcare firms/ industries for finding out their respective decisive risk factors.

    • Numerical investigation of single bubble dynamics in liquid sodium pool


      More Details Abstract Fulltext PDF

      The single gas bubble rise dynamics in liquid sodium/sodium-potassium alloy (NaK) pool due to entrainment of argon cover gas/non-condensable fission gas (xenon) have received considerable attention in the safe operation of Sodium-cooled Fast Reactor (SFR). Numerical simulation of single bubble dynamics in liquid sodium/NaK pool is an essential intermediate step for the evaluation of rise velocity and shape changes, which are of utmost importance in areas of reactor safety concerned with source term evaluation and cover gas purification. The interFoam solver of OpenFOAM package is used to evaluate inert gas bubble rise dynamics in stagnant liquid metal pool of sodium and NaK. The governing equations are discretized and solved using the Volume of Fluid (VOF) based solver available in OpenFOAM with appropriate initial and boundary conditions.The VOF module of the solver is validated against numerical benchmark data and experimental results available in literature. The bubble dynamics in liquid sodium/NaK pool are studied in terms of trajectory, shape and risevelocity for diameters ranging from 10 to 20 mm, domain aspect ratios and for different gas-liquid systems. The study shows that the bubble rise velocity increases with diameter for liquid sodium systems. The rise behavior ofsingle inert gas bubble in liquid water and sodium pool are compared. The study supports the use of air-water system as a simulant for studying bubble dynamics in liquid sodium systems as suggested by other researchers. The study is very useful and forms an intermediate step towards the development of an OpenFOAM basedcomputational framework to analyze heat and mass transfer from single bubble rising in liquid sodium pool for reactor safety studies.

    • A P-hub median network design problem with preventive reliability approach for before and after hub failure


      More Details Abstract Fulltext PDF

      Hubs are vital elements of communication and transportation networks and play an important role in interchanging the flows of information/passenger/goods. For this purpose, designing a highly reliable hub network is very critical, because inefficiency of even a single hub across the network tends to reduce theefficiency of the whole network in transferring the flow appropriately. In this research, a bi-objective mathematical model was designed to study the situations before and after hub failure. Considering reliability, the first objective was to maximize the flow through the network, and the second objective was to prevent wasting the flow due to a possible hub failure. The lexicographic method was used to solve this multi-objective problem with dependent objectives. This method represents an appropriate solution for problems whose objective functionsare of different priorities or depend on one another. Various cases of different sizes were used to evaluate the model in terms of reliability. Since the hub location problem is an NP-Hard problem of commonly large dimensions, a hybrid meta-heuristic algorithm called ‘‘memetic algorithm’’ was used to have it solved. Thealgorithm was a combination of genetic algorithm with simulated annealing algorithm, where simulated annealing algorithm was used for local neighborhood search. Findings indicated that, consideration of the backup hub tends to enhance route reliability, thereby increasing the flow through the network, as compared tothe case with no backup hub.

    • A novel hybrid TOPSIS-PSI approach for material selection in marine applications


      More Details Abstract Fulltext PDF

      Selection of material in engineering design process is a difficult and elusive task due to enormous number of dissimilar materials availability. For effective selection of materials, the designers have to take into account a number of definite qualitative and quantitative criteria. In the same context, this paper proposes ahybrid TOPSIS-PSI approach for effective material selection in marine applications. In this paper, the selection index value has been calculated by using logical combination of PSI and TOPSIS algorithm and these values have been ranked in ascending or descending order. The highest preference selection index value has been taken as the best alternative for the marine application. To prove the effectiveness of the proposed hybrid TOPSIS-PSI algorithm, two practical examples are considered and the result shows that the proposed procedure provides satisfactory results when compared with past literature. Furthermore, hybrid procedure is performed for selection of best wt.% combination among hybrid aluminum nanocomposites for marine applications based on itsphysical, mechanical and corrosive behavior. The result reveals that 9 wt.% and 6 wt.% reinforced hybrid aluminum nanocomposites have optimum combination of all physical, mechanical and corrosion properties, respectively according to hybrid TOPSIS-PSI approach.

    • On the estimation of absolute grating groove density and inter-grating groove density errors of laser pulse compression gratings


      More Details Abstract Fulltext PDF

      In this paper, experimental studies on estimation of absolute groove density of gratings and intergrating groove density errors are reported with typical detector limited accuracies of ±0.23 lines mm⁻¹ and ±0.005 lines mm⁻¹, respectively at groove density of ~1740 lines mm-1 of holographic laser pulse compression gratings. A simple single detector based optical set-up with fixed optical elements to avoid mechanical eccentric errors, if any, due to goniometric movement of a rotatory stage, has been proposed to estimate absolute groove density of gratings. A modified Fizeau or a modified Michelson interferometer based optical set-up has been used to estimate inter-grating groove density errors of gratings. Various gratings from different manufacturers were examined for their absolute groove densities and inter-grating groove density errors.

    • Designing a omnichannel closed loop green supply chain network adapting preferences of rational customers


      More Details Abstract Fulltext PDF

      The manufacturing industry in India started moving towards multi channel supply chain for delivering products to the customer by his preferred channel of purchasing. But firms are far behind to achieve a goal of omnichannel where there exists better interface between various channels which increases their customerrange and customer comfort. Also, today consumer awareness had increased with respect to environmental protection while at the other end the government rules and regulations also force firms to go with adopting sustainability concepts like green and recycling processes in supply chains. This work deals with this major issue faced by the firms in India that how to integrate multiples channels in designing their supply chain network considering economic and environmental sustainability simultaneously. This paper presents a mathematical model for an integrated multi channel closed loop supply chain network problem, the modeling decisions that include the selection of the entity that will fulfill the demands of the omni channel customer during different time periods and the supplier selection process integrated with the production amounts, inventory levels, stockoutsand shipment quantities. The aim of the model is to minimize the total cost incurred to the customer, total cost incurred in running the supply chain and minimize the total pollution emissions from all the transportations of the products between the different stages of the supply chain. It is a mixed integer programming problemconsidered for a leading kitchenware company in southern part of India and solved using IBM CPLEX optimizer. The result shows that how the dynamics of omni channel customer options affects the optimal supply chain structure.

    • A new design method for FIR notch filter using Fractional Derivative and swarm intelligence


      More Details Abstract Fulltext PDF

      In this paper, a new design method for the finite impulse response (FIR) notch filters using fractional derivative (FD) and swarm intelligence technique is presented. The design problem is constructed as a minimization of the magnitude response error w.r.t. filter coefficients. To acquire high accuracy of notch filter,fractional derivative (FD) is evaluated, and the required solution is computed using the Lagrange multiplier method. The fidelity parameters like passband error, notch bandwidth, and maximum passband ripple vary nonlinearly with respect to FD values. Moreover, the tuning of appropriate FD value is computationally expensive.Therefore, modern heuristic methods, known as the constraint factor particle swarm optimization (CFI-PSO), which is inspired by swarm intelligence, is exploited to search the best values of FDs and number of FD required for the optimal solution. After an exhaustive analysis, it is affirmed that the use of two FDs results in 21% reduction in passband error, while notch bandwidth is slightly increased by 2% only. Also, it has been observed that, in the proposed methodology, at the most 66 iterations are required for convergence to optimum solution. To examine the performance of designed notch filter using the proposed method, it has been applied for removal of power line interference from an electrocardiography (ECG) signal, and the improvement in performance isaffirmed.

    • A framework for intelligent question answering system using semantic context-specific document clustering and Wordnet


      More Details Abstract Fulltext PDF

      The question answering system plays an important role in information retrieval field, where the user is in need of getting a precise answer instead of large collections of documents. The aim of this paper is to investigate techniques for improving sentence-based question answering system. To achieve this, a POS-Taggerbased question pattern analysis model is proposed to identify question type based on pattern template for the user-submitted query. Next, the knowledge base is created from a large corpus by clustering the documents by grouping on domain context. The proposed semantic-word-based answer generator model deals with the user query mapping with an appropriate sentence in the knowledge base. By the proposed models, the system reduces the search gap among user queries and answer sentences using Wordnet. It considers word order, overlap, sentence similarity, string distance, unambiguous words and semantic similarity of words. The proposed algorithm evaluates with benchmark datasets such as 20Newsgroup and TREC-9 QA, and proves its efficiency by statistical test for significance.

    • Effect of micro- and elevated gravity condition on the evolution of stratification and self-pressurization in a cryogenic propellant tank


      More Details Abstract Fulltext PDF

      An efficient way of handling and storing cryogenic propellant is required for future space exploration. In rocketry applications, propellants are stored at subcooled conditions in foam-insulated tanks. Any kind of heat infiltration may lead to stratification and self-pressurization of the tank. The supply of warm propellantbeyond the cavitation limit to a turbo-pump is dangerous and hence additional propellant has to be loaded, which affects the payload capacity. The evolution of stratification during lift-off and accelerated conditions and coastphase will be different from those during normal ambient conditions. During lift-off the gravity value can reach up to 6g and microgravity (μg) conditions at the coast phase. Hence, accurate prediction of the state of propellant at all stage is required for the successful mission planning. A multiphase axis-symmetric CFD model is developed, which can simultaneously account for heat transfer from the ambient and heat exchanges within the fluids during different gravity conditions. The results show that the self-pressurisation in microgravity condition is due to phase change rather than thermal stratification. The flow velocity will be maximum during lift-off and accelerated condition. Hence, greater self-pressurisation happens during the initial period and reduction in pressure rise rate is noticed later, which is due to turbulence of the fluid.

    • Modified designs of broadband E-shape microstrip antennas


      More Details Abstract Fulltext PDF

      Wideband designs of rectangular microstrip antennas embedded with three rectangular slots are proposed. These slots modify resonance frequencies of patch’s TM₁₀, TM₀₂ and TM₁₁ mode, to give bandwidth of more than 500 MHz (>50%). Further, pair of slots cut variations of rectangular patch with three slots is proposed. Additional pair of slots tune the next higher order TM₁₂ mode frequency with respect to modified lower order resonant modes, resulting in further increase in bandwidth of more than 600 MHz ([55%). Increased bandwidth obtained in the proposed designs is more than that obtained from equivalent E-shape and U-slot cut patches. As compared with Ψ-shape patch antenna, the proposed configurations yield wider bandwidth with 38% reduction in patch size.

    • The thermodynamic analysis of an electrically supercharged Miller Cycle gasoline engine with early intake valve closing


      More Details Abstract Fulltext PDF

      Electric superchargers are able to improve the thermodynamic process of gasoline engines by selfadapting running state to dominate intake air in all operation conditions. This paper proposes a novel approach for electrically supercharged Miller Cycle with early intake valve closing based on thermodynamics to settle thefuel economy problem of gasoline engines at low load operations by taking advantages of the domination of electric superchargers to intake air. Electrically supercharged Miller Cycle with early intake valve closing was realized by matching an electric supercharger, redesigning the inlet cam, and setting the intake valves closing before bottom dead center while keeping intake valves opening constant. An over-expanded engine cycle is attempted to be used to promote thermal efficiency by increasing geometric compression ratio, in addition to maintaining the effective compression ratio. Here, it has been attempted to systematically analyze the law of the energy losses of the electrically supercharged Miller Cycle gasoline engine with early intake valve closing using a reliable thermodynamic model from the perspective of heat engine. The results indicate that electricallysupercharged Miller Cycle with early intake valve closing could improve the thermal efficiency of gasoline engines by significantly decreasing the pumping and exhaust losses.

    • Longitudinal oscillations of a circular cylinder in a micro-polar fluid: case of resonance


      More Details Abstract Fulltext PDF

      The problem of the longitudinal oscillations of a circular cylinder along its axis of symmetry in an incompressible micro-polar fluid and the flow generated due to these oscillations in the fluid is considered. The Stokes flow is considered by neglecting nonlinear convective terms in the equations of motion on the assumption that the flow is so slow that oscillations’ Reynolds number is less than unity. Here we get a rare, but distinct special case referred to as resonance in which material constants are interrelated in a particular way. In nonresonancecase, all material constants are independent and are not related. The solution in this case cannot be obtained as limiting case of a non-resonance problem. The velocity and micro-rotation components of the flow for the case of resonance and non-resonance are obtained. The skin friction acting on the cylinder is evaluatedand the effect of physical parameters like micro-polarity and couple stress parameter on the skin friction due to oscillations is shown through graphs.

    • Recurrence studies of insect-sized flapping wings in inclined-stroke plane under gusty conditions


      More Details Abstract Fulltext PDF

      Global recurrence plots (GRPs) and windowed recurrence quantification analysis (WRQA) are two recurrence paradigms which find wide applications to detect the onset of instability in a dynamic system. The present work reports the attempt to employ these recurrence paradigms to assess the effect of frontal gust on theforce patterns of an insect-sized flapping wing in the inclined-stroke plane. Horizontal and vertical forces generated by the flapping wing in the presence of gusts of the form uG/UW = u/Uw + (uG/UW) Sin (2π fg /fw t) were numerically estimated in the 2D reference frame for Re = 150. Nine gusts with combinations of the ratio of gust frequency to wing’s flapping frequency, fg/fw = 0.1, 0.5 and 1 and ratio of gust velocity amplitude to root mean square averaged flapping velocity, ug/uw = 0.1, 0.5 and 1 were considered. Recurrence studies of the forces were carried out to find out the gusty condition, which would trigger an onset of unstable behaviour. Studies indicated a possible onset of instability in the force patterns for gust with fg/fw = 0.1 and ug/uw = 1. The onset of unstable behaviour was prominently captured by WRQA of the vertical force coefficient based on determinism (DET) and laminarity (LAM) series.

    • Unsteady hydromagnetic flow and heat transfer of a viscous fluid near a suddenly accelerated flat surface

      S DHOLEY

      More Details Abstract Fulltext PDF

      The unsteady boundary layer flow and heat transfer near a suddenly accelerated flat surface in an unbounded mass of hydromagnetic viscous fluid with the combined influence of the magnetic field, viscous dissipation, internal heat generation/absorption and thermal radiation have been investigated. A new similarity transformation is recommended, which transforms the hydromagnetic boundary layer equations into a set of non-linear ordinary differential equations. These equations are then solved numerically using the finite-differencemethod for some values of the governing parameters that involve unsteadiness parameter β heat source/sink parameter λ Eckert number E, magnetic interaction parameter M, radiation parameter N and Prandtl number Pr. The influence of these parameters on the velocity as well as the temperature field is investigated in detail. In particular, the impact of viscous dissipation (measured through E), which is a strong function of constant reference temperature Tref , on the temperature field has been investigated in different situations. This entails considering the value of Tref as positive or negative depending on whether the surface temperature is higher or lower than the fluid temperature. The analysis reveals that there exists a critical value of E depending upon the values of the other physical parameters for which the surface heat flux vanishes. Below this critical value, heat flows either from the surface to the fluid or from the fluid to the surface depending on whether Tref > or <0.

    • A Russell-based model for estimating overall and divisional efficiency in two-stage production systems with sets of convex hulls in intermediate products


      More Details Abstract Fulltext PDF

      Data Envelopment Analysis (DEA) research works have been recently examining production systems with a two-stage network structure. These studies consider system operations that are performed in two stages where intermediate products play dual roles: they are the outputs of the previous stage and inputs to thenext stage. This dual role is incompatible with Pareto–Koopmans dominance in activity analysis. Also, disregarding intermediate products in assessing performance of two-stage systems compromises the models. Thepresent work introduces a new production possibility set for two-stage network production systems by considering a convex hull for intermediate products. In addition, new models are introduced for evaluating overall efficiency and divisional efficiency of production units from the Pareto–Koopmans efficiency perspective. The proposed models are developed based on an enhanced Russell graph model for efficiency evaluation of two stage network production structures with convex hulls for intermediate products. Numerical examples are further provided for illustration purposes.

    • Arc characteristics of GTAW under high pressure


      More Details Abstract Fulltext PDF

      The characteristics of GTA (gas tungsten arc) under high pressure has been studied in this paper to reveal the special arc phenomenon from the ambient pressure. A totally new hyperbaric GTAW experiment system has been built, including a hyperbaric environment simulator (0.1-3.5 MPa), an automatic GTAWmachine with cooling, the moving equipment, a camera with filter and a central control unit. The arc morphology, the electric static characteristic, the arc voltage and arc length are studied with the increment of ambient pressure from 0.1 to 3.5 MPa, respectively. The relationship between the ambient pressure and the arcvoltage is deduced according to the balance of heat-input and heat-output. The deduced function also verified the results from the welding experiment.

    • Design and application of a novel snubber cell for soft switched pwm dc–dc converters


      More Details Abstract Fulltext PDF

      This paper introduces a novel active snubber cell for soft switching PWM DC–DC converters. In the proposed converter, the main switch is turned on under zero voltage transition (ZVT) and turned off under zero voltage switching (ZVS). The auxiliary switch and all of the other semiconductors in the converter areturned on and off with soft switching (SS). There is no extra voltage stress on the semiconductor devices. Besides, the proposed converter has simple structure and ease of control due to common ground. In this treatise, the theoretical and mathematical analysis of the proposed converter are presented and the design procedure of converters is also provided. The simulation study at 100 kHz switching frequency and 600 W output power are conducted. The experimental prototype of converter is operated under same conditions. Both the simulation andthe experimental prototype exhibits similar performances. Ultimately, the efficiency of the proposed converter is 95.7% at nominal power.

    • Measurement of productivity changes for general network production systems with stochastic data


      More Details Abstract Fulltext PDF

      Performance evaluation of network production systems has been widely studied in recent Data Envelopment Analysis (DEA) literature where internal relations of sub-units are taken into consideration. Most of prior work assumes network systems to have simple series or parallel structures. Complexities of some practical production processes require development of DEA models for their effective analysis. However; input, intermediate products and/or output data are often stochastic and linked to exogenous random variables in most applications. The current study extends Malmquist Productivity Index (MPI) for investigating productivitychanges of general network production units with stochastic data in a DEA framework. The proposed stochastic performance analysis models are then transformed into deterministic equivalent non-linear forms so they couldbe simplified to deterministic programming with quadratic constraints. Numerical examples including an application to productivity evaluation of branches of a university system are presented to illustrate the applicability of the proposed framework

    • Microstructure and mechanical properties of autobody steel joined by friction stir spot welding


      More Details Abstract Fulltext PDF

      This work investigates the microstructural features and mechanical properties of joints of lowcarbon steel friction stir spot welded by changing tool rotational speed and dwell time. Microstructural evaluation, peeling, tension-shear and micro-hardness tests were employed to characterize the properties of the welded joints. The results show that an increase in the tool rotational speed promotes the formation of more weld flash, and it equally changes the flash morphology from ring flash to a combination of ring and serrated flashes. A fine immediate tool contact region is formed in the stir zone (SZ) and the width of the ITCR increases with the tool rotational speed and tensile-shear failure load. Besides the ITCR, the SZ uniquely consists of the shoulder undersurface region (SZ I) and vortex region (SZ II) with average grain sizes of about 10.9 and 16.66 lm, respectively. An increase in tool rotational speed improves the micro-hardness, and the average diameter of the weld nugget (from 4.5 to 5.5 mm). Base metal–nugget interface failure (650 rpm) and base metal neck inginduced failure (1250 rpm) are the two forms of weld failure obtained in the welded samples.

    • Data integrity verification: a novel approach for cloud computing


      More Details Abstract Fulltext PDF

      Data integrity verification (DIV) is one of the primary security checks for data stored on cloud servers. Cloud users have assured the safety of their data with frequent checking of data integrity. In this paper, we provide an efficient and effective DIV approach that offers effective and privacy-preserving audit structure.The main building components of our approach are a multi-power variant of the Paillier cryptography system with homomorphic tag. The Paillier cryptography system along with homomorphic tag assigns a unique and verifiable value to each data block, which helps performing dynamic data operations in cloud environment. To demonstrate our approach, we have implemented an application on Hadoop and MapReduce framework. We have tested this application with respect to the different parameters. The efficiency of the proposed method hasbeen shown through the experimental results. Our method has shown a good improvement over those of the other modern systems.

© 2017-2019 Indian Academy of Sciences, Bengaluru.