KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 10 of 11

Some of the metrics are blocked by your
consent settings
2D-CNN Based Classification on Water Leakage Identification in Automatic Pump
(2026-01-01)
Satthamsakul, Sutham
;
Rattanakun, Kritsana
;
Khummongkol, Rojanee
;
Tangsrirat, Worapong
This study proposes an approach for classification of unusual in automatic water pump systems, especially on a water leakage, one of the key issues that can significantly affect system efficiency and cause serious damage. The method utilizes motor current signals, which are transformed into 2D spectrograms using the Short-Time Fourier Transform (STFT). These spectrograms are then classified by a 2D Convolutional Neural Network (2D-CNN) Designed to determine between usual and unusual operating conditions with high accuracy. Experimental results indicate that the model can effectively detect unusual and leakage events when trained with appropriate parameters, such as a learning rate of 0.001 and 60 training epochs. This model serves as an efficient tool for preventing system failures and reducing maintenance costs in automatic water pump systems.
Some of the metrics are blocked by your
consent settings
Development of a Capacitive Sensing Module to Detect the Concentration of Isopropyl Alcohol for Covid-19 Disinfection Using Small Capacitance to DC Converter
(2024-01-01)
Sriratana, Witsarut
;
Faijaree, Manasnit
;
Sriratana, Lerdlekha
;
Satthamsakul, Sutham
This paper presents the development of a sensing module for analyzing the concentration of disinfectant alcohol commercially available in Thailand during the Covid-19 epidemic situation to ensure the ingredients and percentage concentration by focusing on simplicity, low cost with reliable performance. Under the basic physical structure of the capacitance change [1]-[5], it consists of a New Sensor Module developed to fully interface with the small capacitance to DC converter for capacitive sensor [2] model. Inside the sensing structure, there were 2 modules of metal conductors. Each conductor was 3 mm. thick, placed in parallel with a fixed distance between the module of 1 mm. The cross-sectional areas of the plates of those two modules were different in order to compare the efficiency of the detector developed. The cross-sectional areas of the plates of Module A and B were 4 cm<sup>2</sup> and 1 cm<sup>2</sup>, respectively. Experiments were performed with 3 concentrations of alcohol, 65%, 70% and 75%, to enhance the concentration detecting performance of the sensing module developed. The signal conditioning was generated from a small capacitance to DC converter for capacitive detection with a frequency of 100 kHz, which was the most suitable frequency for exciting the capacitance change which was the capacitive reactance of the alcohol used in this study. From experiments, it was found that the sensing module with a small cross-sectional conductive material would provide more accurate results in terms of linearity and repeatability. The errors were less than 2.72% compared to the module with a large cross-sectional conductor material which would have an error up to 4.73%.
Some of the metrics are blocked by your
consent settings
Contamination Measurement Techniques in Hydraulic Fluid Based on Hall Effect Sensor
(2024-01-01)
Satthamsakul, Sutham
;
Khummongkol, Rojanee
This paper proposed two methods to examine metal particles in fluid using Hall Effect Sensor under the condition of nondestructive testing based on two conditions: test of particles in oil samples, and test of particles in industrial machinery simulation. Hall Effect Sensor was a low-cost device had ability to detect magnetic fields in objects, so this work applied Hall Probe to measure the density of metal in liquid. Hereby, hydraulic liquid was selected to examine because it was one of the most use in industrial machines. By using the sensor, two types of experiment were designed: test of oil sample, and test of simulation process, where the results showed that Hall Effect Sensor had ability to measure particles in both two types of experiment. For oil sample test, the sensor was able to detect metallic powder when the number of contaminants were greater than or equal to 10 milligrams (equivalented to NAS 15 standard). On the other hand, to measure particle in simulation process, magnetic field were generated in two ways, i.e. use of electromagnet, and use of permanent magnet. For using electromagnet as magnetic field source, the result showed that temperature had an influence on metal particle evaluation. Therefore, to get the most accurate value, system's temperature had to employed in calculation process. While, the result of using permanent magnet to generate magnetic field showed that the relationship between output voltage and amount of metal particles were probably in linear form. To apply this method in industrial applications could be an inexpensive way to help officer in detecting metal particle in oil which was able to extend service time of hydraulic liquid properly.
Some of the metrics are blocked by your
consent settings
Fish Recognition Optimization in Various Backgrounds Using Landmarking Technique and YOLOv4
(2022-01-01)
Kuswantori, Ari
;
Suesut, Taweepol
;
Tangsrirat, Worapong
;
Satthamsakul, Sutham
The identification and categorization of fish is a popular and fascinating research topic. Many researchers have developed expertise in fish detection, both underwater and outside the water, which is particularly beneficial for population management and aquaculture. This paper proposes a fish recognition approach using the landmarking methodology with YOLO version 4 to identify and categorize fish with different backdrop circumstances. The approach can be used both underwater and on land. The proposed approach was evaluated using four distinct types of fish from the BYU dataset. The final test result determined that the accuracy reached 96.60%, with an average classification score of 99.67% at the 60% threshold. The result is 4.94 % better than the most frequent traditional labelling approach.
Some of the metrics are blocked by your
consent settings
Mathematical Model for an Autonomous Vehicle to Fulfill the Closed Area
(2022-01-01)
Namngam, Kampanat
;
Chitsakul, Pakkinee
;
Satthamsakul, Sutham
This research presents the mathematical model to roam an autonomous vehicle in a closed region with minor coincident areas. This model is Blind Algorithm, no literature that concern on mathematical model. It is essential for agricultural technology, for example, shoveling soil for planting, spraying liquid fertilizer, harvesting agricultural products, to control the movement of the vehicle with the redundant path is as less as possible is benefited to decrease the working hours and also the burning oil and more for cost-saving without any expensive hardware. The almost wholly filled farmland depends on the shape, the size of the area, and the vehicle dimension.
Some of the metrics are blocked by your
consent settings
Analysis of electric current by using MySQL database on web server for machine performance evaluation: A case study of air conditioning system
(2020-10-13)
Sriratana, Witsarut
;
Khagwian, Vittaya
;
Satthamsakul, Sutham
This study developed a performance analyzing system for industrial machines by using an air conditioning system as a case study. Electricity consumption of the air conditioner compressor operating normally under controlled temperature in the room was gathered by applying the principle of quantitative data collection. Work efficiency of the system was analyzed by using the big data pattern. Due to sufficient information, it can further contribute to the application of AI (Artificial Intelligence) in analysis of the failures or causes of machine malfunctions. Data collecting system in this study was designed to analyze electric current and various parameters including power voltage, temperature and humidity, real power and power factor in order to recheck as well as increase the reliability of the system. All collected data were stored in the MySQL database on a web server and managed by phpMyAdmin that could allow users to access the information chronologically on daily, monthly, or yearly basis via a computer screen. The test results showed that the designed system was accurate with a small variation about 0.07 A or 1.78% compared to the standard digital ammeter. For power voltage, an average difference compared to the standard digital voltmeter was approximately 0.4 V or about 0.17%.
Some of the metrics are blocked by your
consent settings
Development of compatible induction coil with pure AC for hall effect sensor: A study on metal materials thickness
(2018-12-10)
Sriratana, Witsarut
;
Khwankityotha, Yannadech
;
Satthamsakul, Sutham
This study presents an analysis of the parameters to generate the electromagnetic fields of the induction coils for the application of Hall Effect sensor to classify the aluminum thickness and the classification of the metal materials. In this research three types of induction coil are considered: Type A with one layer, Type B with two layers, and Type C with three layers. Four experiments were conducted. The electromagnetic field of Type C is found to be 11.90% greater than Type B, and 29.26% greater than Type A. The core temperature of Type C with 1 A current applied was found to be 3.50% hotter than Type B and 5.14% hotter than Type A. The classification of the aluminum thickness with the frequency range of 100 Hz to 1.4 kHz are suitable for identifying the thickness of 1, 2 and 3 mm. In terms of the classification of metal materials with all the three types of the induction coil, it can be separated clearly from the samples which consist of steel, stainless steel and aluminum. The total error by using Hall Effect sensor is ±0.5702%.
Some of the metrics are blocked by your
consent settings
Development of sensor module for detecting the magnetron power based on microwave oven frequency
(2018-12-10)
Satthamsakul, Sutham
;
Sriratana, Witsarut
This study presents the design of sensor module to be used in analyzing the electromagnetic power created from magnetron by using microwave as the case study on basis of Rectenna (Rectifying antenna) consisting of 2 main parts including sensing part and transduction part. On part of sensing part, the design would be made on the basic principle of monopole antenna; on part of transduction part, the design would be made on the basis of rectifier circuit, to function in converting the electromagnetic power received from magnetron to be in form of DC current, here the Hall Effect Sensor that was installed inside the current sensor module was used as the measuring equipment, with Arduino UNO R3 as the controller for recording electrical current including temperature. The experimental results would be used to analyze the appropriate form of sensor module. This research compared sensor module for 3 forms including Type A, Type B, and Type C, with different sizes of diameter of wires consisting, by comparing the efficiency in preventing the leak of measured power. According to the test of sensor modules in 3 forms on the same condition, it was found that the means of electrical current from converting the electromagnetic power of microwave.
Some of the metrics are blocked by your
consent settings
Test of oil lubricant viscosity by using hall effect sensor
(2018-08-13)
Sriratana, Witsarut
;
Nakklinkoon, Sethawut
;
Satthamsakul, Sutham
This paper presents the test of oil lubricant. The viscosity of lubricant would decrease and the replacement of lubricant is required when the lubricant is used for a period of time. Hence, viscosity is a major factor that can indicate the quality of the lubricant. The fundamental method known as falling ball viscometer that is basically from Archimedes' principle including Newton's laws of motion were applied for the test. The sphere shaped object was replaced by a standard disc with the drill holes. The experiment was conducted by four new lubricant samples: SAE10W-30, SAE15W-40, SAE20W-50 and SAE90. From the reference position, the standard disc was released to the lubricant container and a timer was counted simultaneously by the microcontroller. The timer was stopped when Hall Effect sensor detected the disc at the bottom of the container. Falling time was compared to the viscosity from the standard rotational Viscometer, namely Brookfield Viscometer model DV-II+Pro to assess the relationship equation between falling time and viscosity. From study, the average error of falling time measurement was 0.109 Sec or 2.106% and the average viscosity error from the equation was 6.171 cP or 3.089% at room temperature (27°C). In addition, comparing to the standard Viscometer, the average percentage error was about 2.113%. Even though the accuracy of this method would not be outstanding, it is efficient with cost advantage compared to the expensive standard Viscometer.
Some of the metrics are blocked by your
consent settings
Measurement of metal particles in oil lubricant using hall effect sensor under temperature conditions
(2017-11-10)
Inthawatkul, Itti
;
Sriratana, Witsarut
;
Satthamsakul, Sutham
This paper develops a Hall-Effect sensor based system to assess the performance of a lubricant. The measurement methodology is based on the fact that the electromagnetically induced voltage is directly proportional to the amount of contaminants, which in turn affects the state of the lubricant. Two main cases are considered: the metal particles contaminant, and viscosity. The sample lubricant viscosity was experimentally found to be 49.2874 Cst, which is well within the range of internationally accepted values. Furthermore, the lubricant flow rate was also found to be temperature dependent. The output Hall voltage for the contaminated lubricant was 0.8 A for liquid flow rates between 0 to 30 l/h, which significantly deviates from the non-contaminated lubricant. These results suggests that the presented approach is a simple, low cost system capable of accurately assessing the performance of lubricants without consulting the manufacturer's information, even under normal operational loading.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results