Search

Time plays a pivotal rule in successful control of sepsis-associated infections and mortality in ICU patients. Accurate and confident prediction of sepsis ahead of time is a challenging task owing to the complex nature of the underlying pathophysiological mechanisms. Missing data in the laboratory values which have been reported by medical findings to be important for forecasting sepsis makes the problem worse. To address the problem of missing data, last observation carried forward (LOCF) method has been adopted in this paper. This has been proved to be a simple yet effective idea. The remaining missing features were substituted with zero, which is one type of constant imputation. Out of the 40 available features, 19 were selected by performing the analysis of variance (ANOVA) test. The class imbalance problem was addressed via cost-sensitive learning which gave more weight to the minority class data points based on their proportion. Finally, a random forest classifier was trained upon the data. The model obtained a utility score of approximately 57% and 22% respectively on the two datasets which were taken from Physionet/Computing in Cardiology (CinC) challenge, 2019.

Sepsis is a dangerous condition that happens when the body’s reaction to contamination results in tissue harm, failure of organs, or sometimes even death. Early recognition and anti-infection sepsis-treatment are basic to enhancing sepsis results, where every hour of deferred treatment has been related with an escalation in mortality. Along these lines, this image gives a reasonable view as to the significance of early discovery of sepsis. This study focuses on detecting the onset of sepsis based on 40 features from ICU patients using the logistic regression classifier. The heavy class imbalance between septic and nonseptic patients is addressed by giving more weight to the minority class. Pearson correlation coefficient was used to select the features useful to predict the target variable and the data was standardized afterwards. Using the scoring system provided by PhysioNet/Computing in Cardiology (CinC) Challenge 2019 ,the model achieved a score of 32.4% on the training dataset (20336 subjects) and 14% on the unseen test dataset (20000 subjects).

The world's electricity demand is ever increasing since the digitalization and industrialization have begun. Bangladesh as a developing country has a power sector that has been one of the fastest-growing ones in South-Asia. A relentless development of its economy is urging the government to increase the generation capacity every year. Bangladesh Power Development Board's master plan aims to generate 39,000 MW by 2030. This study is a representation of the 2030 generation scenarios in terms of real power flows between the 9 electrical zones of Bangladesh. An optimal power flow model with minimum-cost generations is developed in MATPOWER simulation environment. The DCOPF model simulated here, considers a loss-less network where only real power flows among 13 interconnections for the 9 zones are considered. The power flow values and directions obtained here give an approximate idea of the future generation and transmission of real power in Bangladesh for the year 2030.

Nowadays, high availability of optical and inertial sensors in fitness trackers and mobile phone attracts researchers attention on human activity recognition(HAR) through photo-plethysmography and inertial sensor data. It is also more suitable than camera based HAR in different circumstances. In this paper, we propose a wavelet packet transform based feature extraction technique from PPG along with accelerometer and gyroscope data to differentiate various physical activities like high and low resistance biking, running and walking. Wavelet packet transform(WPT) is applied on acquired signals and a number of statistical features and entropy feature are extracted. Random forest(RF) classifier is employed in recognition. A 5-fold cross validation is done to examine the performance and an accuracy of 97.8% is achieved which surpasses existing algorithms worked on the same dataset.

In high performance chalcogenide based thin-film solar cells, a bilayer intrinsic and doped zinc-oxide (ZnO) n-type stacks are used as window layer. Intrinsic ZnO and aluminum-doped ZnO thin-films were deposited on soda lime glass substrates using the standard recipe by radio-frequency (RF) magnetron sputtering with thickness 80 nm and 350 nm respectively. The structural, optical and electrical properties of the as deposited thin-films were studied by various characterization techniques such as X-Ray Diffractometry (XRD), UV-Vis spectroscopy and Hall-effect measurements. It was found from structural studies that both films are crystallographically oriented along (002) and (103) planes with hexagonal Wurtzite type. The optical transmittances and bandgaps were found around 90% (in visible and near infrared regions of spectrum) and 3.28 eV respectively. According to our findings, it is concluded that these as deposited films are suitable as window layer for promising chalcogenide Copper zinc tin sulphide (Cu 2 ZnSnS 4 , CZTS) based thin-film solar cells.

CZTS (Copper Zinc Tin Sulfide) is a propitious absorber material for photovoltaic applications and gaining popularity for its material abundances and non-toxicity. In order to get good adhesion with substrate and at the same time low film resistivity, bi-layer Molybdenum (Mo) thin-films are frequently used as back electrode in substrate-structured CZTS based Thin-film Solar Cells (TFSCs). In a bi-layer structured Mo film, usually a thin high-pressure (HP) bottom layer serves to give better adhesion and a thick low-pressure (LP) top layer shown lower film resistivity. In this work, structural study of two different stacks (total 1μm film thickness) of HP layer and LP layer for both as-deposited and Vacuum Thermal Annealing (VTA) @560°Cfor 45 minutes were performed. Crystals were more oriented along [200] plane instead of [110] plane and oxidation at the Mo surface were observed for thicker bottom layer. It has been found that, for thinner bottom layer, not only increased crystallite size (from 19.27 nm to 21.26 nm for as-deposited and VTA treated case respectively), but also compressive stressed Mo films were found. It is evident that this bi-layer Mo thin-film is suitable for CZTS solar cell fabrication.

Frequency adjustment and protection of a wind turbine due to wind gust is a salient aspect of the grid-tied wind farm. Both of them require conserving system stability. This paper presents an investigation of the converter based frequency adjusting method and wind turbine protection. A grid-tied wind power system model using the Doubly Fed Induction Generator in MATLAB is used. To regulate the active power output, synchronization between wind velocity and wind turbine speed is adjusted. Thus, wind farm operated at grid frequency and maximize turbine output. The control strategy of the converter based frequency synchronization of a grid-tied wind farm also includes protection subsystem. The protection system executed by receiving information from the logical block implemented in the wind farm model. It provides wind turbine protection by terminating wind turbine from a grid in abnormal wind conditions. The Simulation result validates the results and control methods.

Continues expanding of wind power harnessing demands an effective control system to improve quality power delivered from wind turbine to grid. A stable power system fully depends on the synchronization that means matching of frequency between the source of power generation and existing power grid. The prime objective of a wind farm controller is to ensure effective operation of the wind farm in any wind speed condition. Frequency deviation between the wind system and grid can challenge the well-built operation of a power system and can imbalance the synchronization between wind farm and grid. The strategy of frequency adjustment is developing time to time which starts from using droop controller based Induction generator to converter based Double-fed Induction generator. This paper summarizes an analytical study on various frequency adjustment mechanisms for a grid-connected wind farm.

Highest consistency of transmission line is very noteworthy due to enhancement of demand. Voltage stability and steady state limit are matter of great concentration as it plays an imperative function in Bangladesh Power System (BPS). In this paper we have used Modal analysis technique to scrutinize steady state voltage vulnerability and Continuation Power Flow (CPF) technique to assess steady state voltage stability limit in East-West Interconnector System (EWIS) of BPS using NEPLAN software. Inadequate reactive power in the power system causes voltage collapse which is mainly a local problem. This can be overcome by adding shunt compensative device like Static Synchronous Compensator (STATCOM). Inclusion of shunt FACTS (Flexible AC Transmission System) increases the stability and loading factor. The comparative result with and without STATCOM is also shown in BPS.

In this paper a comparative study has been performed between combined Thyristor controlled series capacitor (TCSC), Static VAR compensator (SVC) and combined TCSC, Static synchronous compensator (STATCOM) to enhance the steady state, dynamic state and transient state stability of East-West interconnected system (EWIS) of Bangladesh Power System (BPS). Two TCSCs are connected in series with both interconnectors and two SVCs and STATCOMs are connected alternatively in shunt mode at the receiving end of both interconnectors. Steady state, Modal and Transient stability analyses are performed to show the improving effect of series shunt combined compensation on stability and capacity enhancement of EWIS. In case of shunt compensation superiority of STATCOM is also shown in comparison to SVC. In both the cases of comparison series compensation level is kept identical for fair comparison.

This paper presents a modified design of a sliding mode controller based on fuzzy logic for a dc-dc boost converter. Here a proportional - integral (PI)-type current mode control is employed and a sliding mode controller is designed utilizing fuzzy algorithm. Sliding mode controller ensures robustness against all variations and fuzzy logic helps to reduce chattering phenomenon introduced by sliding controller, thereby increasing efficiency and reducing error, voltage and current ripples. The proposed system is simulated using MATLAB / SIMULINK. This model is tested against variation of input and reference voltages and found to perform better than conventional sliding mode controller.

In conventional grid-connected PV system, the control of real and reactive power is undertaken adjusting both modulation index and phase angle of the inverter. The maximum power point of the PV array is tracked by adjusting duty cycle of the dc-dc converter or modulation index of the inverter. The PWM pattern has to be optimized each time the index of modulation is changed which involves excessive computational complexities and time. In this paper, a new phase-angle controlled grid-connected PV system is simulated in Matlab where power flow is controlled adjusting the phase angle of the PV inverter keeping the modulation index constant and thereby maximum power from the PV array is extracted. Since modulation index is kept constant, PWM pattern is optimized only once and used for any phase angle value. This involves lesser computations and ensures optimized PWM pattern for inverter control that makes the system more efficient. Modelling, analysis and simulation results have been presented to show the effectiveness and simplicity of the proposed technique.

Researchers proposed several techniques to reduce the harmonics present in the input current of 12-pulse rectifier. But the transformers used in the rectifier contain more low order harmonics in their input currents. In this paper a novel active filtering method for input current wave shaping of the transformers used in a 12-pulse rectifier is proposed. With the proposed method both the rectifier and transformers are shown to draw near sinusoidal input currents from the utility. This active filtering technique uses high switching frequency PWM Boost converter that shapes the input currents by eliminating high frequency harmonic components with the aid of small size LC filters. The size of the LC filters is reduced to a considerable extent with the proposed technique. Detailed design and simulation results are presented to show the effectiveness of the proposed method.

In a power system of conventional parallel connected generators, share of real power and reactive var of an incoming generator are controlled by adjusting shaft power input and field excitation. The scenario of load sharing by a grid connected PV system is however different since no prime mover or excitation source are present. Due to serious power crisis, there are needs for evacuation of PV power to grid systems. However, this needs intensive analysis on the load sharing phenomena. In this paper, aspects of load sharing of a grid connected PV system are analyzed, and a strategy is proposed where the load sharing task can be undertaken controlling both modulation index and phase angle of the inverter. It is seen that both real power and reactive var are affected upon change in index of modulation and phase angle of the PV inverter. Analysis and simulation results are presented to demonstrate effectiveness of the proposed control technique.