Gongcheng Kexue Yu Jishu/Advanced Engineering Science

Abstract : Improving the operational performance of wastewater treatment plants can be effectively approached by means of model simulation. GPS-X model was used in this study. Calibration and validation of the model were carried out, with various sensitive parameters subject to modification, with results found within the prescribed parameters for R and RMSE. Sensitivity analysis then indicated that the most important factor for reducing nitrogen and phosphorous concentrations was the readily biodegradable fraction; thus, the IR, RAS ratio, DO, and WAS flows were reduced from 3% to 1%, from 100 to 20%, from 3.5 to 2 mg/L, and from 3,500 to 1,000 m3/d, respectively, producing an optimization that saved 688.4 Kw.h in energy and gave a sludge reduction of 32%. These results showed that an IR percentage of 3% is not appropriate. Decreased rbCOD thus necessitates a chemical upgrade, which was implemented in this case by means of adding an external carbon source, represented by acetic acid, propionic acid, methanol, and glycerol, with good results. These additions led to improvements in terms of reduced TN and TP by suitable ratios. The best external carbon source was thus determined to be methanol, while glycerol was less effective than the others. The process of pre-denitrification was compared with the post-denitrification process by means of the addition of methanol as an external carbon source, which gave good results for the reduction of TN in the post-denitrification process, by up to 80%; however, the effect on other pollutants was to increase concentrations..
Full article

Abstract : We introduce the new class quasi -(ζ,η)- normal operators. Composition operators, Weighted composition operators, Composite multiplication operators of Quasi –(ζ,η)- normal and their adjoints on L^2 (ℷ) are described..
Full article

Abstract : Present a newborn class of an operator (℘≀,ϱ)-Paranormal and (℘≀,ϱ)-*-Paranormal Composition, Weighted Composition and Composite Multiplication operators on L^2 space are characterized. Also characterized (℘,≀,ϱ)-paranormal and (℘,≀,ϱ)-*-paranormal operators on Hardy space..
Full article

Abstract : In today's modern world, the Internet is highly crucial. A larger number of wired and wireless sensors characterize the Internet of Things (IoT). These sensors are networked devices that create massive amounts of data, reaction time, latency, and security vulnerabilities. These problems create delay in decision making in some situation which is very extreme in case of Healthcare, Smart vehicles, Industry 4.0 etc., the data moving from the sensor devices have to reach the cloud and from there it has to reach back to receivers. Therefore, in order deal with such huge data we use fog computing, a well-known distributed architecture. Fog computing works with the aim to enhance the processing, intelligence, and accumulation of data closer to the Edge devices. The proposed framework helps in reduce latency as we place a Fog node device between the cloud and the edge device where data is generated and sent to the cloud and retrieved from the cloud. The framework is designed in such way that the different sensor devices can be placed on it and collects the sensor data from them. To test the framework functioning, without the fognode and with the fognode, comparing the latency with packet transfer rate from sensor devices to the cloud and vice versa. In this paper, we are considering two different case studies to test the proposed functioning of the framework with the following case studies i) Related to medical are like diabetes and cardiovascular illnesses are considered for prediction based on patient health records and ii) Vehicle theft where the owner get an alert of the theft of the vehicle within few seconds of the theft. In Medical case study initially, patient health data is acquired and stored using Sensor devices. A group of patient health records is first subjected to the unique rule-based clustering technique. Finally, we consider diabetes and cardiovascular diseases in our study. To test the performance of the suggested task, a rigorous experiment and study using healthcare data were conducted. Various Machine Learning (ML) algorithms are applied to patient health information, and the prediction result, as well as the accuracy of each algorithm, is generated. The findings show that the proposed work accurately predicts cardiac and diabetes problems, identifies which algorithm has the highest accuracy among them, and reduces burden on cloud. Devices having inbuilt sensors used in this work are pulse oximeter, smart watch, glucometer and thermometer. In Vehicle theft case study, there are numerous anti-theft solutions for smart vehicles based on GSM and cloud computing. The previous models have excessive latency and bandwidth. Fog computing is a creative new concept that combines existing solutions with GPS and GSM. IoT development for a fog- based vehicle anti-theft technology that uses a mobile application to pinpoint the exact position of a stolen vehicle. Low latency, low bandwidth, fast responsiveness, and high security are all advantages of fog computing. Thus with the use of the above proposed framework can reduce the latency and bandwidth in these applications and helps to avoid the delay decision making. Further this framework can be used to analyze and improve the performance and consumption of power in smart cities, smart homes and smart appliances..
Full article

Abstract : The use of credit cards has grown significantly as the globe moves quickly toward digitization and cashless transactions. Additionally, there have been more fraud-related operations, which costs financial institutions a great deal of money. As a result, we must examine and distinguish between fraudulent and legitimate transactions. In this study, we wanted to implement the comprehensive model training procedure from beginning to end. As a result, we will have the best possible model to categorize the transaction into normal and aberrant varieties. Machine learning methods have been used in the fight against credit card fraud; however, fraud detection systems have yet to prove particularly effective. The relatively breakthrough of deep learning has been used in various fields to address complex challenges. In this study, we investigate several models for machine learning to spot fraudulent credit card activity. We compare the results produced by each model as well as their performance. The SMOTE methodology yields the most successful outcomes. It has been argued that under-sampling the majority class (the normal class) could effectively improve a classifier's sensitivity to the minority class. This research shows that, compared to just undersampling the majority class, the procedure we chose for oversampling the lower (abnormal) standard and undersampling the upper (normal) standard can enhance the analyzer's conduct..
Full article