TNI Journal of Engineering and Technology

A Social Media Monitoring System for Thai University

Tanatorn Tanantong, Sasitorn Kreangkriwanich, Supranee Thaskhwan — Tue, 24 Dec 2019 13:57:28 +0700

A Social Media Monitoring System aims to gather comments on Twitters referring to Thai Universities. Examples of 20 public universities and 20 private universities, which are selected based on the World University Ranking of Webometrics, are used for analyzing and visualizing in different formats, e.g., ranking mentions of universities and grouping frequencies of high uses or common words in sentences. It demonstrates that universities are referred in which topics in each time duration. This system is suitable for those who want to employ the analyzed data for making a decision in management and improvement of university efficiency. For example, an analysis of comments concerning university activities from students that whether comments are rapidly considered.

A Study of Beacons System Comparison with RFID Time Attendance System for Human Resource Management

Yuranan Musor — Tue, 24 Dec 2019 14:05:30 +0700

Time Attendance System is considered as the information technology of human resource management which is important for human resource personnel in summarizing the collection of working hours for payroll. Nowadays, technology is commonly used to record the time attendance in most organizations is using an RFID, magnetic stripe employee cards tap on the card reader to record time. From the study, it was found that Beacon technology can be applied as an employee data reader by using Mobile application of employees to connect and transmit data via Bluetooth signal to for time recording of departure of employees. This research aims to study the use of beacons in the time attendance system. And comparing the results of recording data between beacons and RFID in the time attendance system by the timer used to record system information. From the individual timer trial results, the RFID time recording system can time faster than the average of Beacons, 3.4 seconds, but in total timer Beacons Can time faster than the average of RFID time recording system 13.33 seconds and is satisfied by the test user more than RFID time recording system.

Application of Taguchi Method for Defect Reduce Assembly Process on Printed Circuit Board A Case Study of An Electronics Part Factory

Somporn Vongpeang, Anyarat Prasanjai — Tue, 24 Dec 2019 14:30:41 +0700

This research aimed to reduce a defect in printed circuit board assembly process. The amount of defect of the production process at a case study company was 4.39 percent before the improvement. The steps of improvement were based on the Pareto chart and cause – effect diagram to analyze the problem and design the experiment by Taguchi method in order to find the suitable factors. The effective factors in the process were a squeegee Speed, a squeegee force, and a board size height. The experiment was designed as a tool to determine the appropriate set of parameters in order to find the signal to noise ratio that had a smaller the better value.
The result of this research showed that the appropriate parameters to set up the machine were: a squeegee speed was 30 millimeters per second, a squeegee force was 20 Newton, and a board size height was 1.0 millimeters. After the experiment, it showed that defect reduced to 2.58 percent from 4.39 percent.

Artificial Neural Network Modeling for Prediction of Tensile Strength in Pressure Vessel Welding with Steel ASTM A537 Class 1

Prachya Peasura — Tue, 24 Dec 2019 13:44:18 +0700

This research described to the determine a mathematic model using artificial neural network (ANN) for predicting the of tensile strength in the pressure vessel steel ASTM A537 Class1 grade with submerge arc welding process. The following welding parameters were studied: the welding current, voltage and travel speed. The resulting welding samples were examined using tensile strength tests which were observed microstructure with scanning electron microscopy (SEM) and determine a suitable mathematic model. The research results reveal that using a neural network model with the proposed mathematical model, which tensile strength represents 3 neurons for the input 14 neurons and 1 output neurons (3-14-1). The Levenberg-Marquart training algorithm was also train for weight and bias network. The neuron of log-sigmoid for input layer, tan-sigmoid for hidden layer and purelin for output layer activation function was assigned. The mean square error (MSE) and coefficient of determination (R2) for tensile strength predict was showed that of 0.284. The welding conditions which have the highest tensile strength was showed that microstructure phenomenon could be that fine pearlite and spheroidized pearlite with distribute intensity in heat affected zone. The mathematical model that can be effectively applied to predict and quality control of welds to obtain tensile strength according to the standard requirements, which was reduce production costs and increase work efficiency.

Design and Implement of Two Axes Automatic Sun Tracking with Fuzzy Controller

Surachat Chantarachit — Tue, 24 Dec 2019 16:19:30 +0700

The solar energy is widely used in Thailand. Most of all are fixed on the roof or ground. To increase efficiency of harvesting energy, two axes solar tracking is designed and constructed. The position of the sun is detected by using four LDR sensor. Those sensors are divided by partition in order to know the light direction of the sun. Those data from the sensor are sent to micro-controller to determine the motion by fuzzy controller. The rule of controller is designed to be concerning of energy consumption. Two linear actuator is applied to rotate solar panel. Both of actuator is decoupled. Fifty watts of solar panel is applied to collect the energy. The solar tracking of this research can gain more energy than fixed panel 14 percent. The experimental results are presented and compared.

Determination of Inventory Replenishment Factors with Lateral Transhipment in a Multiple Warehouses and Multiple Retailers Distribution System

Anchalee Supithak — Tue, 24 Dec 2019 15:00:12 +0700

The objectives of this research are to study the lateral transhipment of multiple warehouses and multiple retailers distribution system and to analyse the effects of related factors composing of demand dispersion and filling rate. The developed model is simulated by varying the demands at each location and the filling rate of demands at each location to observe the expected number of stockout in the system. The ten replications of each scenario are simulated and the results show that the total number of item’s stockout increases when demand dispersion is high. Retailer with the lowest demand in a distribution results in the highest number of units’ stockout. By varying the filling rate, the distribution with low dispersion demand among retailers trends to has lowest number of units stockout.

Enhanced MIMO-SCMA Detector by Deep Learning

Triratana Metkarunchit — Tue, 24 Dec 2019 15:09:52 +0700

In this paper, we propose a deep learning approach for multiple input multiple output - sparse code multiple access (MIMO-SCMA) signal detection by using a deep neural network via spreading the procedure of the message passing algorithm (MPA). The MPA can be transformed into a sparsely connected neural network. The neural network can be trained off-line and then implemented for online detection. Besides when the neural network has been trained, the network weights corresponding to the edges of a factor graph. From the simulation result of the MIMO-SCMA system over the quasi-static Rayleigh fading channel, found that the neural network detection better performance than the traditional MPA.

Glauco Dtex Eye Glasses: Eaylr Diagnosis Device for Glaucoma Detection

Paween Wongkornchaovalit, Thunyaseth Sethaput — Tue, 24 Dec 2019 15:19:47 +0700

Glaucoma is the second cause of vision loss and blindness, not only in developing countries deal with it, but developed countries also face it. The main problem of blindness from glaucoma is a lack of recognition, to clarify, the patients do not know whether they have glaucoma or not. As time pass by, it continues to develop from time to time, and most of the patients will realize about it when it has developed into a severe stage. According to the ophthalmology research, if the area ratio between the optic cup to the optic disc is greater than 1/3, patients will be classified as having glaucoma. The aim of this research is to develop the device to capture the image of the optic disc. Then, the image provided by the prototype, with the aid of image processing algorithm based on the area ratio of the optic cup to the optic disc, can be used to classify glaucomatous patients from healthy individual. From the result, the captured image of the optic disc, the location of the optic nerve head is clear. The image of the optic disc apparently shows brighter and dimmer section, which represents the optic cup and optic disc respectively. The image can be used for analysis of opened angle glaucoma via image processing algorithm, which are provided by many researchers [ 1,19,21,22].

Investigation of Energy Conservation Potential for Convenience Stores in Thailand

Tue, 24 Dec 2019 15:34:09 +0700

The convenience stores in Thailand presents its continuing growth in both urban and rural areas. The stores possess various electric equipment for refrigeration, air-conditioning, lighting, cooking, with 7/24 service; resulting in high electricity demand. Those stores have been persistently conducted energy conservation; however, there is still a great potential for energy consumption reduction.

This research conducted detailed energy auditing to investigate the energy usage and its influencing factors. Nine convenience stores participated in the study. The results showed that the stores consumed electricity at 173,284 kWh/year, average peak power 20 kW, energy intensity 823 kWh/m².year. The refrigeration shares 59–64% of the total energy consumption. The air-conditioning came the second with the consumption of 26-28% while the electric lighting consumed about 5%. This study proposed eight energy conservation measures including (1) Reflective ceramics painted roof, (2) Insulated roof for heat resistance, (3) Solar photovoltaic rooftop, (4) Reduction of glazed area, (5) inverter air-conditioner, (6) Inverter compressor for refrigerated open showcase, (7) Inverter compressor for walk-in cold room, and (8) 3-layers glass for walk-in cold room.

A mathematical model was developed for the store to present the energy usage in each system and the whole shop. For Thailand, the estimated conserved energy in the standalone and the commercial building’s store, implementing particular measure and combined measures, from 2019 – 2027 is about 9.9%. Therefore, these measures should be further evaluated for their investment worthiness.

Improving the Performance of Tree Algorithm using Mixed Splitting Algorithm

Warakorn Srichavengsup — Tue, 24 Dec 2019 15:41:53 +0700

The aim of this research is to enhance the performance of tree algorithm using mixed splitting algorithm. Binary tree is very efficient when the number of users is small. In case of large number of users, Ternary tree performs better than other algorithms. Therefore, this paper introduces mixed splitting algorithm. The details of this algorithm are as follows. When the number of the collided users is small, this algorithm will split the collided users into 2 groups. In case of large number of users, this algorithm will split the collided users into 3 groups. From the results, we found that the probability of success of the proposed algorithm tends to be higher than the Binary and Ternary algorithms. With this reason, the proposed algorithm performs superior performance than other algorithms.

SARIMA Models Applications for Demand Forecasting and Supply Planning of Natural Rubber in Thailand

Pawinee Paisri, Weerapat Sessomboon, Rojanee Homchalee — Tue, 24 Dec 2019 16:04:23 +0700

Rubber is the important economic crops and export products of Thailand. Studying and forecasting the demand of rubber could be used as information for supply planning including production and cultivate area. Time series analysis was used to formulate the forecasting model for export volume of Concentrated Latex, RSS Grade 1, and Crepe Rubber. The appropriate forecasting time series models were and , respectively. After that historical statistics of production and stock were used to estimated rubber production and cultivate area corresponding to forecast including domestic demand and export. In the year 2022, Thailand should produce rubber about 1.573 million tons, which requiring cultivate area is about 6.561 million rai, respectively. The research results could be used to plan and set the appropriate policy for Thai rubber production. Consequently, Thailand should reduce the rubber plantation area by 12.659 million rai and then promote the planting of other economic crops. This is to achieve a balance between demand and supply.

Unfolding to the Past: Temporal and Dimensional Perception Enrichment for Image Impression Improvement

Kulwadee Somboonviwat, Thitiporn Lertrusdachakul — Tue, 24 Dec 2019 16:10:38 +0700

An image is in essence a record of visible objects at a particular point in time. This paper proposes a method based on adaptively integrating depth cues with relative motion for enriching temporal and dimensional perception of images. The key idea of our method is to construct a natural-looking animated sequence of images from the original one by introducing differing depth cues and varying time frames to individual generated images. The differing depth cues are created by utilizing variations of visual details and color saturation, whereas the presentation time duration of each image is determined based on structural similarity of consecutive images. The resulting animated image can give a sense of increasing depth as well as a temporal dynamics from past to present. The proposed method can be applied to image impression improvement, especially in the exhibitions or presentations of historical places and artifacts.