Volume 5

ABSTRACT:Load forecasting is a challenging task in the setting of modern power systems, which have risen in complexity as conventional and non-conventional energy sources have been integrated into an increasingly varied energy environment. Utility companies are under growing pressure to not just provide cost-effective and adequate power generation, but also to maintain system dependability for today’s discriminating customers. While there are several load forecasting systems, neural network-based techniques appear as a potential alternative due to their ability to reveal hidden subtleties within the input/output load data connection, resulting in fewer predicting mistakes. Artificial neural networks (ANNs)-based short-term load prediction methods have become more widely used, successfully overcoming issues related to weather, temperature, humidity, precipitation, air pressure, and the shifting patterns of human and industrial activity. This has made accurate load forecasting easier. We present FITNET, a novel feedforward neural network model designed for short-term load prediction (STLF), as our contribution to this effort. FITNET is Zunique in that it can adjust to events occurring in real time and allows training with a wide range of input kinds and sequences. We collected data from the ISO New England NE-Pool region over a period of four and a half years and combined it into a single, coherent dataset. Important inputs include time-related components and meteorological characteristics, such as day and night, dew point, and dry bulb temperature, with weekdays having a substantial impact on the output data. To improve the performance of the ANN model, we carefully examined alternate neuron configurations, using the Levenberg-Marquardt backpropagation approach for training. Extensive testing of our suggested model across both weekly and daily load forecasting methodologies continually shows outstanding efficiency, with the ANN model constantly having a forecasting MAPE of less than 1%. This finding emphasizes the model’s stability and its potential to considerably improve the dependability and cost-effectiveness of power generation in today’s complex and ever-changing energy landscape.

ABSTRACT: The major breakdown of Covid-19 was held in the year 2020 around the world. It is the quickest spread disease found in the world. Its symptoms involve cough, temperature, flu, muscle aches, headache, and many others. This study finds the top five clinical symptoms that would lead to COVID-19 in any person and evaluates with supervised learning classifiers: Support Vector Machine (SVM), Gaussian Naïve Bayes, Logistic Regression, K-Nearest Neighbor (KNN), and voting (ensemble) were used. For the evaluation of this, a dataset from Kaggle was selected with 5326 instances and 21 features. Precision, recall, and F-score are the selected performance measures. Different machine-learning classifiers were applied to find the core symptoms of Covid-19. As a result, cough, fever, breathing problems, attending gatherings, and traveling were the prominent symptoms found in this study.

Abstract: In the present study, the effects of laser irradiation on microstructure, morphology, and hardness of lithium discilicate (Li₂Si₂O₅), commercially known as e.max Press are examined and the findings are simulated by using MATLAB. The polished discs were irradiated by Nd: YAG pulsed laser with 100 and 150 laser shots and the one disc was kept unexposed. XRD analysis confirmed the change in phase after laser irradiation however a decrease in dislocation line density was observed. The scanning electron micrographs showed that laser irradiation initially caused the upper surface to melt resulting in a rough surface. Further increase in laser shots produced cracks within the sample. Vickers hardness was employed to observe the hardness of non-irradiated and irradiated samples. An insignificant decrease in hardness (0.33%) was observed when the material was irradiated by 100 laser shots. The irradiated surface exhibited more roughness for bonding as compared to the unirradiated surface which could be beneficial in handling e.max Press during repair in direct dental applications. Moreover, the fuzzy analysis of the data would help in modern digital dentistry.

Abstract: Protein leakage through urine, at a level exceeding 3.5g per 1.7m^2 of body area per day, is a symptom characteristic of Nephrotic syndrome. It’s a diverse condition with symptoms including edema, hypoalbuminemia (≤ 25g/L), albumin excretion greater than 40 mg/m^2 per hour, uPCR ≥ 2,000 mg/g or 3+ protein in urine test. Springtime is when this syndrome is more common. A detailed literature review was conducted by using credible databases and scientific journals. This review concludes that in nephrotic syndrome, which is caused by metabolic and genetic defects, patients have to face proteinuria, edema, hyperlipidemia, and hypoalbuminemia. This study additionally focuses on the use of corticosteroids to treat nephrotic syndrome and the proper doses needed for effective therapy. Our study identified the complications related to nephrotic syndrome including the syndrome itself and post-treatment complications. Corticosteroids have been reported to side effect and relapse of nephrotic syndrome has also been identified as a health issue.

ABSTRACT: Helium plasma has been widely used for the production of active species by mixing in other gases which are used for surface modification of metals. Helium plasma is generated using 100 Hz pulsed DC source and its characterization is carried out by using optical emission spectroscopy. The spectra of helium is recorded at filling pressure (0.5-3.0 mbar), source power (25-150 W) and inter-electrode distance (3-5 cm) using Ocean Optics HR 4000 spectrometer. It is found that production of active species of helium strongly depends on discharge parameters. Evolution of the selected emission intensities of He-I and He-II are presented in this paper. The emission intensities of He-I (501.3 nm and 667.7 nm) as a function of above parameters are used for the determination of electron temperature. The spectroscopic technique based on the measurement of relative intensities of two spectral lines of the same atom is used to evaluate the electron temperature, which is found to vary from 0.82 eV to 1.89 eV depending on the various discharge parameters.