This project aims to propose a design for a red offal processing room in a pork processing plant that processes 500 pigs per day or 80 pigs per hour. Each pig weighs approximately 105 kilograms, with 3.47% of the weight consisting of red offal. The process involves separating liver, gall bladder, heart, lungs, spleen, and kidneys as required. These parts are then chilled in cold water to reduce their temperature to below 7°C before packaging and sealing. Sorting is based on the number of pieces and weight, depending on the type of product. The processing times of sorting chilling and packaging vary depending on the product's type and size. The design was developed using data collected from the current production line and referenced standards. The room layout was planned using Systematic Layout Planning (SLP) principles to analyze activity relationships within the room and define functional areas. Equipment sizes and the required number of operators were calculated to ensure optimal use of space. The red offal processing room was designed with an area of 56 square meters. After the layout design was completed, a 3D model was created using SketchUp 2024, and the workflow and operations were simulated and analyzed using Flexsim 2024
บริษัท เบทาโกร จำกัด (มหาชน) ได้มีแผนการขยายโรงงานแปรรูปสุกรทั้งในประเทศและต่างประเทศ การออกแบบกระบวนการผลิตจึงเป็นสิ่งสำคัญที่ทำให้โรงงานที่จะก่อตั้งนั้นสามารถใช้งานได้อย่างมีประสิทธิภาพทั้งในด้านกระบวนการการผลิตและความสะอาด เพื่อทำให้สินค้าที่ส่งมอบแก่ผู้บริโภคมีคุณภาพที่ดี โดยผลิตภัณฑ์ที่สำคัญของโรงงานแปรรูปสุกร ได้แก่ กลุ่มชิ้นส่วนสุกรตัดแต่ง และกลุ่มเครื่องในสุกร ซึ่งประกอบไปด้วยเครื่องในแดง และเครื่องในขาว โครงงานนี้ได้รับมอบหมายให้ศึกษากระบวนการทำงานเพื่อออกแบบห้องเครื่องในแดงให้รองรับการผลิตที่สูงขึ้นและมีความหลากหลายของผลิตภัณฑ์มากขึ้น โดยกระบวนการทำงานภายในห้องถูกต้องเป็นไปตามหลักการปฏิบัติที่ดีสำหรับโรงแปรรูปสุกรของกรมปศุสัตว์
คณะวิศวกรรมศาสตร์
This cooperative education project aims to enhance the efficiency of Hydrogen Manufacturing Unit 2 (HMU-2) and Pressure Swing Adsorption 3 (PSA-3) by using AVEVA Pro/II process modeling and a Machine Learning model for process simulation. The study found that the AVEVA Pro/II model predicted outcomes with deviations ranging from 0–35%, including a hydrogen flow rate deviation from the PSA unit of 12%, exceeding the company’s acceptable limit of 10%. To address this, a Machine Learning model based on the Random Forest algorithm was developed with hyperparameter tuning. The Machine Learning model demonstrated high accuracy, achieving Mean Squared Errors (MSE) of 8.48 and 0.18 for process and laboratory data, respectively, and R-squared values of 0.98 and 0.88 for the same datasets. It outperformed the AVEVA Pro/II model in predicting all variables and reduced the hydrogen flow rate deviation to 4.75% and 1.35% for production rates of 180 and 220 tons per day, respectively. Optimization using the model provided recommendations for process adjustments, increasing hydrogen production by 7.8 tons per day and generating an additional annual profit of 850,966.23 Baht.
คณะสถาปัตยกรรม ศิลปะและการออกแบบ
A Photographic series that expresses the abstract states of myself, towards the question of existence that results from being surrounded by expectations of both surrender and freedom of expression, this series focuses on my own subjectivities in order to bring back memories of almost forgotten feelings and make them clear once more.
คณะวิทยาศาสตร์
This special problem aims to study and compare the performance of predicting the air quality index (AQI) using five ensemble machine learning methods: random forest, XGBoost, CatBoost, stacking ensemble of random forest and XGBoost, and stacking ensemble of random forest, SVR, and MLP. The study uses a dataset from the Central Pollution Control Board of India (CPCB), which includes fifteen pollutants and nine meteorological variables collected between January, 2021 and December, 2023. In this study, there were 1,024,920 records. The performance is measured using three methods: root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination. The study found that the random forest and XGBoost stacking ensemble had the best performance measures among the three methods, with the minimum RMSE of 0.1040, the minimum MAE of 0.0675, and the maximum of 0.8128. SHAP-based model interpretation method for five machine learning methods. All methods reached the same conclusion: the two variables that most significantly impacted the global prediction were PM2.5 and PM10, respectively.