Asian Journal of Convergence in Technology
Not a member yet
868 research outputs found
Sort by
Development of Solar PV Charge Controller System for Rural Application
In recent year many intelligent techniques and approaches have been introduced in photovoltaic (PV) system for the utilisation of free harvest renewable energy. The modern day power system is undergoing rapid and dynamic metamorphosis from the legacy system, in the direction of an intelligent power system. The use of renewable energy resources as rural electrification at the distribution level is on the rise alongside advances in the efficiency of associated technologies and automation of the power sector. This paper presents a cost effective solar charge controller (SCC) using programmable intelligent controller (PIC) to control and coordinate the functions properly. The source code for the PIC microcontroller will be written in C to obtain accurate and efficient disconnecting or reconnecting action automatically. Accordingly, battery can be disconnected from solar cell when overcharging and reconnected while under discharging. The proposed charge controller is equipped with Liquid Crystal Display (LCD) to display the battery charging status and amount of current flown from solar cell to load via microcontroller. The Details of design for the construction of SCC using crystal oscillator, optocoupler, Dusk-Down sensor and an In-build digital voltmeter, are presented. The construction, operation and experimental result of our proposed smart solar charge controller indicates that it is more cost effective, reliable and functions properly
SIMULATION AND ANALYSIS OF DGA ANALYSIS FOR POWER TRANSFORMER USING ADVANCED CONTROL METHODS
In any power system or grid network transformer device use for power from one circuit to another circuit without changing the frequency with high efficiency levels. Due to number of usages Transformer protection is very important now a days for electric supply which is fault free, efficiency and to increase the transformer life cycle. This paper is a brief discussion about the concentration of different gases like CO, CO2, H2, C2H6, C2H4, C2H2 and CH4 related faults which is known as DGA analysis with the help of various classical techniques gives different conditions for the same sample unit. This paper presents MATLAB simulation of ANN and Machine learning based high accuracy design techniques for DGA analysis and the results are compared with the classical techniques like Key Gas Method, IEC Ratio method, Duval triangle Method and Rogers Ratio Method, but in this paper we have done only Duval triangle Method for Comparison using Matlab. The Simulation result of the proposed methods shows that overall DGA analysis using Machine learning algorithm is better than conventional Duval triangle method Performance
An IoT based Real-Time Monitoring of Water Quality System
With the emerging technologies and new generation, each field is developing gradually, like IoT, Artificial Intelligence, Big Data, Cloud Computing, etc. Currently, the Internet of Things (IoT) has become the pillar of all connected technologies, without which we can’t imagine our single day. IoT is useful in every aspect of our life. Also, water is an essential component for all living beings, and monitoring of water quality and its parameter is a must to get safe water. The proposed model, water quality monitoring in Aquaculture based on IoT which is used to determine the different parameters of water like, pH value, turbidity, temperature, to detect the bad quality which may lead to any disease or adverse effect to a living being. So, it is crucial to detect the quality of water to have safe water for a healthy life. Traditional approaches include a manual collection of samples with complex methodology and ineffective because of time-consuming, high cost, lack of real-time monitoring and testing is done later in the laboratory
STUDENT ATTENDANCE SYSTEM BASED ON THE FACE RECOGNITION
Face Recognition Attendance system is important in schools and institutions. In school and institution is not easy to maintain the attendance records in paper and file. Therefore, an existing system that uses a face recognition attendance system increases accuracy and it takes less time as compared to another method. There are a lot of systems available such as face recognition using “IR motion” sensors, etc. With automatic face recognition attendance system, hardware devices(software), helpfull. But the challenge is to keep all the nerves properly without getting hurt. the face recognition attendance system we use javascripts,node.js and library are used tensorflow.js and face-API. It can take pictures from 50- 70cm. We create a graphical interface that takes pictures, builds a database ,and trains the database with one click
Application Of Machine Learning Techniques For Fake Customer Review Detection
Now-a-days with the increasing demand of the web , online marketing is additionally becoming progressively popular. This is often because; tons of products and services are easily available online. Hence, reviews of these products and services are vital for customers also as sellers. But, to gain profit or promotion, scammers produce fake reviews. These fake reviews written by scammers prevent customers and sellers reaching actual opinion about the products. Hence, fake reviews or spam reviews must be detected and eliminated so as to prevent misleading potential customers. In our work, supervised and semi supervised learning techniques are applied to detect spam review
Water Supply System for a Rural Village in Malaysia
The aim of this research was to design and build a water supply system for a village in rural Malaysia. Very often technologies previously developed for water supply are not known or specifically upgraded to suit a particular population (or village). This has resulted in 600 million humans still living in abject poverty. Companies can be called in to develop water supply systems but too often their profit motive will make the project inviably expensive. This project was accomplished with a very tight research grant, lots of trial and error and engineering knowledge. The village is located 103 kilometers from Kuching. The final system is a 3Ø, 4 HP pump placed on a floatable wooden barge near the river pushing water up 120m over an elevation of 9m to a Pump House located behind the Community Hall of the village. Another 3Ø, 1 HP pump is used to pump water from the 9000L holding tank to the 38 individual homes of the village. Upon achieving a supply of water, other aspects of developing the village using modern technology was worked upon. This will enable villagers to get out of the poverty trap and approach life with renewed confidence
An Overview of Different Techniques to Provide Semantics to Remote Sensing Data
Due to the rapid advancements in the remote sensing field, there is an immense amount of data being generated. This data is raw. Hence, it requires semantics and one way of providing these semantics is land use and land cover classification. We present different techniques that can be used for the same. After analyzing the different classification models, it was observed that the machine learning models performed poorly as compared to the deep learning models. VGG19 gave the best accuracy of 97.64%. In order to provide semantics to remote sensing data, the different classification models are essential and this work can be further extended into diverse domains
System to provide Reading Aid to Visually Impaired People
Artificial intelligence and Machine learning is fast evolving in today’s world running towards digitization and automation, ranging from cancer detection using scans in Healthcare to driverless automatic electric cars. The system to provide reading aid is an approach to help said differently abled people to gain emotional independency in life, when it comes to reading text visuals, be it in the form of a book, newspaper or notice board. It will help increase scope of education for visually impaired which was formerly relatively less possible due to unavailability of books in Braille. For this technology, OCR (Optical Character Recognition) and Signal Processing (Raw images) is use, coupled with a camera module attached to the Raspberry Pi
Online E-commerce Venture - Choose The Best
This paper entitled 'E-commerce venture -choose the best' plays a key role in the upliftment of digitization and automation of small-scale businesses. The majority of the small businesses in India don't have a website according to the survey conducted by GoDaddy-world's largest technology services provider. In short, the proposed system aims at building online e-commerce ventures to satisfy the needs of small businesses and enterprises by providing extensive machine learning features
Mobility Management in Vehicular Adhoc Network based on RSA Algorithm
Vehicular Ad-hoc Network (VANET) is become very popular in academia as well as in industry. The important characteristics of VANET faces difficulties on network issues. Specifically the mobility management will helpful to provide a seamless connectivity and improve the Quality of service. So we present an overview of mobility management in VANET based on RSA algorithm. During vehicle to vehicle communication driver do not want to share his personal information like vehicle names, license, number plate, speed position, moving port routes and user information. To protect all the information and save from any external attacks such algorithm is effective. Devised work is also useful and helpful to find the cause of accident or any liabilities. Recent work on mobility management is discussed in this paper. Proposed work includes an ID based crypto system for valid user. For this paper we are implementing RSA algorithm. In which data is encrypted and then the data is converted to cipher text and then decrypted with a private key. And the receiver will get original data