KMITL

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

Browse

Search Results

Now showing 1 - 10 of 28

Some of the metrics are blocked by your
consent settings
Hybrid machine learning models: A comprehensive, data-driven evaluation with diverse data partitioning strategies for net radiation estimation
(2025-04-30)
Bajao, Kristian Lorenz
;
Phetpan, Kittisak
;
Chophuk, Ponlawat
;
Suwalak, Rattapong
Surface net radiation (Rn) is crucial for climate modeling and agricultural management but is often not readily available, especially in regions like Thailand. Accurate prediction of Rn is essential for estimating evapotranspiration, which is vital for irrigation planning and agricultural productivity. This study develops a hybrid machine learning framework that incorporates K-Nearest Neighbors (KNN) for missing data imputation, Random Forest-Recursive Feature Elimination (RF-RFE) for feature selection, and machine learning models (Multi-layer Perceptron, K-Nearest Neighbors, and Random Forest) for prediction. The research evaluates various data partitioning methods, including hold-out split, K-fold cross-validation, and growing-window forward-validation (gwFV), alongside hyperparameter tuning using GridSearch to enhance model robustness and prevent overfitting. The primary objectives are to develop and evaluate the hybrid ML models for daily Rn estimation using basic meteorological inputs (temperature, relative humidity, and sunshine duration), assess the impact of different input combinations on prediction accuracy in Sawi, Chumphon, Thailand, and compare data partitioning techniques to determine the optimal model performance. Utilizing FAO56PM-calculated Rn as a reference, this study finds that the Random Forest model, with average temperature and sunshine duration (M2) as inputs evaluated under the gwFV method, achieves the highest stability and high accuracy (R² of 0.972, RMSE of 0.457 MJ m-2 day-1, and MAPE of 3.50%). The Random Forest demonstrates strong generalization capabilities, making it a reliable choice. Even models using only sunshine duration (M3) perform adequately, offering a solution when data availability is scarce. This study concludes that hybrid machine learning models, combined with careful data partitioning, significantly improve Rn estimation. These advancements provide valuable insights for climate modeling, agricultural management, and irrigation scheduling, particularly in data-scarce regions.
Some of the metrics are blocked by your
consent settings
Dual-Band Band-Stop Filter using Multiple Hexagonal Microstrip Line for Chipless RFID Sensor
(2023-01-01)
Suwalak, Rattapong
;
Lertsakwimarn, Kittima
;
Lertwiriyaprapa, Titipong
;
Phongcharoenpanich, Chuwong
;
Torrungrueng, Danai
The dual-band band-stop filter using a hexagonal-shape microstrip line is presented in this paper. The chipless RFID with the band-stop filter used to determine the relative permittivity of material under test and identify characteristic based on the unique responded signal from the chipless RFID sensor. This paper proposed the multiple microstrip line structure with a quarter-wavelength transformer to generate an identification (ID) code, i.e., ID-01, ID-10, and ID-11 at the center frequency of 2.2 GHz and 2.7 GHz. From the results, the proposed chipless RFID sensor can be identified and determined the relative permittivity of the LWC.
Some of the metrics are blocked by your
consent settings
Flexible Chipless RFID Sensor for permittivity sensing of Cylindrical Impedance Surface
(2023-01-01)
Suwalak, Rattapong
;
Phongcharoenpanich, Chuwong
;
Torrungrueng, Danai
;
Lertwiriyaprapa, Titipong
This paper presents the chipless RFID sensor characteristics when placed on the cylindrical impedance surface. The chipless RFID sensor is designed based on the multiple resonant printed on the flexible substrate of Polyimide (r = 3.5 and tan δ = 0.0027). The humidity of the material under test is identified using the chipless RFID sensor based on the multi-resonators. This chipless sensor can be generated 5 bits of identification (ID). Furthermore, the orientation effect of the plane wave is studied. The results shows that the flexible chipless RFID sensor can be identified the different dielectric constant that related with the water content in the material. Therefore, this sensor can be applied to recognize the humidity state of material under test.
Some of the metrics are blocked by your
consent settings
Face Detection Approach to Classify Emotions Based on Facial Expression in Depressive Disorder
(2023-01-01)
Suwalak, Rattapong
;
Sukkaeo, Tuksina
;
Promwanrat, Thanawut
;
Satjawiso, Satjalinee
;
Werachattawan, Nisan
The Multi-Task Cascaded Convolution Neural Network (MTCNN) is presented in this paper to classify the emotion and generate the facial dots as a representative of the patient. In depressive disorder diagnosis, the facial expressions can be used to observe the behavior of the patient. From the results, the system can be classified the emotion into 5-class i.e., happy, angry, disgusted, neutral, and surprised. For emotions of happy, angry, neutral, and surprised, the accuracy is more than 98 %, and for disgusted emotion is 96 %. Furthermore, the system can generate the real-time facial dots for emotion classification. Therefore, it can be a candidate to apply to collect and analyze the emotions of the patient under the privacy policy in a depressive disorder.
Some of the metrics are blocked by your
consent settings
Machine learning-based prediction of nutritional status in oil palm leaves using proximal multispectral images
(2022-07-01)
Chungcharoen, Thatchapol
;
Donis-Gonzalez, Irwin
;
Phetpan, Kittisak
;
Udompetaikul, Vasu
;
Sirisomboon, Panmanas
This study evaluated the application of proximal multispectral images accompanied by 4 machine learning approaches for estimating the nutritional status of oil palm leaves. The image responded for five bands: blue, green, red, red edge, and near-infrared regions with a center wavelength of 475, 560, 668, 717, and 840 nm. Average and standard deviation (SD) values from the leaf pixels of each band were extracted, obtaining 5 average and 5 SD values from 5 bands. Thirty-four vegetation variables were generated based on those average and SD values. In total, forty-four variables consisted of 10 average-and SD-based features, and 34 vegetation variables were used as the input candidates for analyses against 10 target variables: nitrogen (N), phosphorus (P), potassium (K), calcium (Ca), magnesium (Mg), iron (Fe), manganese (Mn), zinc (Zn), boron (B), and chlorophyll (SPAD). No significant input came out for modeling with P and Zn based on the stepwise selection. Therefore, 8 nutritional models were proposed in this study. A training set with 50 samples was used to be modeled for each target, and a test set with 15 samples was employed to evaluate the models' performances. Based on random forest (RF), support vector regression (SVR), partial least square regression (PLSR), and artificial neuron network (ANN) applied to be modeled, the models for chlorophyll, N, and Ca predictions were acceptable for screening, and those for K and Mg predictions were acceptable for rough screening. The chlorophyll model developed based on the RF had the predictive statistics in terms of coefficient of determination for prediction (r2), root mean square error of prediction (RMSEP), and standard error of prediction (SEP) of 0.752, 5.46 SPAD, and 5.65 SPAD, respectively. The other 2 screening models developed based on SVR and RF for N and Ca, respectively, gave the performances with the r2, RMSEP, and SEP ranging from 0.655 to 0.718, 0.12 to 0.17%, and 0.12 to 0.18%, respectively. In the case of the 2 rough screening models established using the RF algorithm, the predictive statistics ranged from 0.496 to 0.530 for the r2 and 0.07–0.16% for both RMSEP and SEP. In this study, the Fe, Mn, and B models had poor results presenting the range of r2, RMSEP, and SEP of 0.308–0.491, 2.39–72.9 ppm, and 2.45–62.8 ppm, respectively. Based on the results, this study confirmed that the proximal multispectral information of oil palm leaves had enough significance to account for the status of chlorophyll and macro-nutrients: N, K, Ca, and Mg in the leaves.
Some of the metrics are blocked by your
consent settings
Dual-Band Band-Stop Filter for Chipless RFID Sensor in a Dielectric Constant Determination
(2022-01-01)
Suwalak, Rattapong
;
Phaebua, Kittisak
;
Lertwiriyaprapa, Titipong
;
Phongcharoenpanich, Chuwong
;
Torrungrueng, Danai
This paper presents the dual-band band-stop filter to operate with the chipless RFID tag acts as the RFID sensor for a dielectric constant determination and the identification characteristic based on the signature responded signal from the chipless RFID sensor system. The proposed dual-band filter can generate the dual-stop band of the frequency of 2.2 GHz and 2.7 GHz. The 3-bits identification (ID) i.e., ID-01, ID-10, and ID-11 are obtained from the proposed filter. In addition, the chipless RFID sensor with band-stop filter technique can be determined the dielectric constant of RT/Duroid5880, Fr4 (Glass Epoxy), and low-Temperature cofired ceramic (LTCC).
Some of the metrics are blocked by your
consent settings
A Compact Label-Typed L-Shaped Slot UHF RFID Tag Antenna Mountable on Metallic Cans
(2022-01-01)
Lertsakwimarn, Kittima
;
Suwalak, Rattapong
This paper presents the design of a compact label-type UHF RFID tag antenna which designed for mounted on metallic items in supply chain goods such as metallic cans. The proposed tag antenna is a single copper line with L-shaped slot. This design is intended to obtain a narrow structure to make it possible to attach the tag antenna in the longitudinal and transverse along the metallic cans without being bulky. The tag antenna impedance is easy for turning impedance matching with RFID IC chip by adjusting the length and width of the line. The overall dimension of the proposed tag antenna on the metal actual cans is 94 × 6 × 0.8 m3. The proposed tag antenna covers the targeted UHF RFID bands (920-925 MHz) with -10 dB|Γ| bandwidth of 25 MHz.
Some of the metrics are blocked by your
consent settings
SUBTAWEE Transportation Management on Mobile Application
(2022-01-01)
Maliwan, Nattaphorn
;
Maliwan, Chonlada
;
Suwalak, Rattapong
This paper provides information about the design of the mobile application for transportation management. Initially, the user must be logged in to access the mobile application. It is divided into 3 access roles, i.e., 1) administrator can use job summary, time verification function, and driver function, 2) staff can use the driver's time and time verification function, and 3) driver can use the job function. The application's movement data since login. Confirmation of the time of the car leaving the plant. Job verification on behalf of the driver, including the driver confirming the customer submission, is stored in the entire database, which is notified every time it is logged in and when the work is sent to the customer via the Line Notification, as well as the daily revenue amount and monthly income is displayed. Displays within the proposed mobile application to keep drivers informed. The mobile application is convenient to use and work very well for the transportation management of the SUBTAWEE Concrete Company Langsuan Co., Ltd. This mobile application can be applied to the transportation management and the others similarly process.
Some of the metrics are blocked by your
consent settings
Chipless RFID sensing for dielectric property of light weight concrete
(2021-05-19)
Suwalak, Rattapong
;
Phaebua, Kittisak
;
Phongcharoenpanich, Chuwong
;
Lertwiriyaprapa, Titipong
;
Pathoumvanh, Somsanouk
This paper presents a passive chipless RFID tag (Electromagnetic sensor) based on the modified printed monopole antenna with a frequency filter technique. The proposed chipless RFID sensor is the wireless sensor technology with the nondestructive technique (NDT) testing. This proposed RFID tag is designed to obtain both determination and identification (ID) properties for a sensor application. To determine the dielectric constant of MUT under test, the circular monopole tag antenna is modified using adding the elliptical shape to achieve the wideband operating frequency. Moreover, this sensor is used the stripline frequency filter technique to obtain the 1-bit ID at the frequency of 2.77 GHz. The CST Microwave Studio simulator program uses to design and optimize the passive chipless RFID parameters. Simulated results show the proposed chipless RFID sensor can identify and determine the dielectric property of the LWC under test.
Some of the metrics are blocked by your
consent settings
Curved meander line resonators for chipless RFID sensors
(2019-09-01)
Suwalak, Rattapong
;
Phongcharoenpanich, Chuwong
;
Akkaraekthalin, Prayoot
;
Torrungrueng, Danai
This paper presents a chipless RFID sensor based on the curved meander line resonator. The sensor is designed and printed on Fr4-substrate with thickness of 0.8 mm. The resonator on the RFID sensor acts as a data bit for identification, where we use only one bit in this paper for illustration. Transmitting and receiving antennas on the RFID sensor are properly designed to sense the dielectric property of the material under test. From the results, it can operate from 1.8-2.5 GHz. The gain is 2.14 dBi at the resonant frequency of 2.28 GHz. Therefore, it can be candidate as a RFID sensor with a nondestructive testing technique.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results