Engineering Journal (Faculty of Engineering, Chulalongkorn University, Bangkok)
Not a member yet
1223 research outputs found
Sort by
Neural Network-based Hybrid Estimator for Estimating Concentration in Ethylene Polymerization Process: An Applicable Approach
Get PDFEstimation of a monomer concentration of an ethylene polymerization process has been a challenging problem due to its highly nonlinear behavior and interaction among state variables. Applying of an extended Kalman filter (EKF) to provide the estimates of the concentration based on measured bed temperatures has usually been prone to errors. Here, alternatively, neural network-based hybrid estimators have been developed and classified into three structures which integrating of either EKF or Kalman filter (KF) to neural network (NN) to provide the estimates. The NNs are integrated to provide the estimates’ error or concentration’s estimates corresponding to individual structure for reducing the estimation error. Simulation results have shown that the hybrid estimators can provide good estimates under nominal condition and disturbance cases. However, in dealing with noises, the NN-KF hybrid estimator gives superior robustness with smooth and accurate estimated values
Strengthening of Shear-Critical RC Columns by High-Strength Steel-Rod Collars
Get PDFThis research investigates the strengthening of shear-dominated reinforced concrete (RC) square columns using the high-strength steel-rod collars, which enhance confinement by steel rods around the column perimeter. This method is less intrusive to the existing building with infilled walls because steel rods can penetrate through the walls with minimal openings (holes) at the location of collars. In this study, one control specimen and two specimens strengthened by steel-rod collars were tested. All specimens were subjected to lateral cyclic loading along with a constant axial load. The difference between the two strengthened specimens were the spacing of steel-rod collars mounted on the columns. The spacing of steel-rod collars was 200 mm in the column specimen SC-200, while the other strengthened specimen, SC-100 has a spacing of 100 mm. The unstrengthened column failed in shear while the strengthened columns failed in flexure. In addition, the strengthened specimens failed at the higher load and ductility. Comparing to the unstrengthened column, the lateral load capacity and ductility ratio of the column SC-200 increase by 18 % and 59%, respectively. While, the lateral load and ductility ratio of SC-100 increase by 16% and 69%, respectively. Furthermore, the finite element models of all column specimens are developed using the OpenSees program. The analysis results are found in a close agreement with the experimental results
Modularization of Ship Engine Room Using Design Structure Matrix (DSM) Based on the Genetic Algorithm
Get PDFRecently, the shipbuilding industry has been able to develop new production methods. These new methods promote design automation in order to produce ships more efficiently. The various production concepts, like block division, modularization and building ships with a standard design are possible solutions for improve production. Engine room design, including the piping system, is a complex process; therefore, modularization of its design is an effective strategy to minimize the complexity of the system. In addition, modularization plays an important role. This process requires a considerable number of man hours. This paper presents a new approach for engine room design based on the modularization concept. The characteristics of the proposed method are as follows: • Attention was paid to all piping systems of ship engine room. The cost and weight of the piping system were considered. • To define an effective module, a design structure matrix was adopted. • In the modularization using DSM, the Genetic algorithm is used to obtain modules by considering some constraints like number of pipe connections and pipe cost. This study discusses the details of the above mentioned methods. In addition, simulation test of design optimization of a several piping systems were carried out to illustrate the design optimization procedure in detail and to verify the effectiveness of the proposed methodology
Finite Element Analysis on Reinforced Concrete Columns Strengthened by ECC Jacketing under Eccentric Compressive Load
Get PDFEngineered cementitious composite (ECC) can be used for strengthening of concrete columns due to its similar structure and suitable connection to normal concrete and its special tension behavior. In this study, to analyse the columns, finite element (FE) method was used after verification by experimental results. Reference column was strengthened by normal concrete and ECC jacketing. The effects of type of jacket material, longitudinal reinforcement, compressive stress and ultimate tensile strain of ECC on variations of eccentric load-bending moment (P-M) interaction curves were investigated. Results showed that the use of ECC instead of normal concrete can increase load carrying capacity of strengthened column, due to tensile strain hardening behavior of this material. It was found that, amount of this increase depends on eccentricity of eccentric load and varying from 0.4-23%. In ECC jacketing, tensile cracks are continuous, but in concrete jacketing, there were discrete cracks and more quantity of damages. Due to higher load carrying capacity and better distribution of tensile cracks in ECC jacketing than normal concrete jacketing, the use of ECC is suitable for strengthening of reinforced concrete columns. Load carrying capacity of columns under concentric load and pure bending moment were calculated by theoretical method and the results were compared with FE
Feature Selection for Text and Image Data Using Differential Evolution with SVM and Naïve Bayes Classifiers
Get PDFClassification problems are increasing in various important applications such as text categorization, images, medical imaging diagnosis and bimolecular analysis etc. due to large amount of attribute set. Feature extraction methods in case of large dataset play an important role to reduce the irrelevant feature and thereby increases the performance of classifier algorithm. There exist various methods based on machine learning for text and image classification. These approaches are utilized for dimensionality reduction which aims to filter less informative and outlier data. Therefore, these approaches provide compact representation and computationally better tractable accuracy. At the same time, these methods can be challenging if the search space is doubled multiple time. To optimize such challenges, a hybrid approach is suggested in this paper. The proposed approach uses differential evolution (DE) for feature selection with naïve bayes (NB) and support vector machine (SVM) classifiers to enhance the performance of selected classifier. The results are verified using text and image data which reflects improved accuracy compared with other conventional techniques. A 25 benchmark datasets (UCI) from different domains are considered to test the proposed algorithms. A comparative study between proposed hybrid classification algorithms are presented in this work. Finally, the experimental result shows that the differential evolution with NB classifier outperforms and produces better estimation of probability terms. The proposed technique in terms of computational time is also feasible
A Multi-Objective Variable Neighborhood Search Algorithm for Precast Production Scheduling
Get PDFIn real life, precast production schedulers face the challenges of creating a reasonable schedule to satisfy multiple conflicting objectives. Practical constraints and objectives encountered in the precast production scheduling problem (PPSP) were addressed, with the goal to minimize makespan and total earliness and tardiness penalties. A multi-objective variable neighborhood search (MOVNS) algorithm was proposed and the performance was tested on 11 problem instances. Ten of these were generated using precast concrete production information taken from the literature. One real industrial problem from a precast concrete company was considered as a case study. Extensive experiments were conducted, and the spread and distance metrics were used to evaluate the quality of the non-dominated solutions set. Statistical analysis demonstrated that the result was statistically convincing. Computational results showed that the proposed MOVNS algorithm was significantly better when compared to the other nine algorithms. Therefore, the proposed MOVNS algorithm was a very competitive method for the considered PPSP
Parameter Tuning of the Autonomous Boat in Fish Farming Industry with Design of Experiment
Get PDFThis research studied the feasibility in applying the flight controller of the unmanned aerial vehicle on a pontoon boat in fish farming industry. A small-scale autonomous pontoon boat has been built, equipped with the open-hardware flight controller, GPS receiver, Inertial Measurement Units (IMU) and the magnetometer.
A 23 factorial design based on Design of Experiment (DOE) is carried out to study the influence of three parameters (turn_rate, turn_angle and damping) on the response variable (peak-to-peak deviation from the desired trajectory). Total 24 experiments have been conducted by setting the desired trajectory as a circle with diameter 20 m. The peak-to-peak deviation in each experiment has been observed.
The analyses revealed that the damping has low interaction with the turn angle and turn rate while there is a stronger interaction between the turn angle and turn rate. And the peak-to-peak deviation of the trajectory tends to decrease when the parameter damping was set to high value. The regression model has been derived and plotted as a response surface. The optimized parameters were selected from the plot and perform the experiment with three replications. The results confirm that parameters tuning has improved the performance of the boat significantly.
With the DOE approach, the impact of two or more parameters on a response, the interaction between parameters can be investigated systematically. This approach is an effective way to tuning the parameters and can be applied to various kinds of the autonomous vehicle.
A 23 factorial design based on Design of Experiment (DOE) is carried out to study the influence of three parameters (turn_rate, turn_angle and damping) on the response variable (peak-to-peak deviation from the desired trajectory). Total 24 experiments have been conducted by setting the desired trajectory as a circle with diameter 20 m. The peak-to-peak deviation in each experiment has been observed.
The analyses revealed that the damping has low interaction with the turn angle and turn rate while there is a stronger interaction between the turn angle and turn rate. And the peak-to-peak deviation of the trajectory tends to decrease when the parameter damping was set to high value. The regression model has been derived and plotted as a response surface. The optimized parameters were selected from the plot and perform the experiment with three replications. The results confirm that parameters tuning has improved the performance of the boat significantly.
With the DOE approach, the impact of two or more parameters on a response, the interaction between parameters can be investigated systematically. This approach is an effective way to tuning the parameters and can be applied to various kinds of the autonomous vehicle
Vibration Modelling and Control Experiments for a Thin-Walled Cylindrical Rotor with Piezo Patch Actuation and Sensing
Get PDFThis paper describes a dynamic model formulation and control experiments concerning the vibration behaviour of a thin-walled cylindrical rotor with internal piezoelectric patch transducers. Model development, validation and controller design procedures were undertaken for an experimental rotordynamic system comprising a tubular steel rotor (length 0.8 m, diameter 0.166 m and wall-thickness 3.06 mm) supported by two radial active magnetic bearings. Analytical solutions for mode shapes and natural frequencies for free vibration were first derived using a shell theory model, and these used to construct a speed-dependent parametric model for the rotor structure, including piezo patch actuators and sensors. The results confirm that the developed shell theory model can accurately capture the rotating frame dynamics and accounts correctly for frequency splitting from Coriolis effects. The model is also shown to be suitable for active controller design and optimization. Model-based H2 feedback control using the rotor-mounted actuators and sensors is shown to achieve vibration suppression of targeted flexural modes, both with and without rotation
The Combination Effects of Age-Related Bone Mechanical Property, Cortical Bone Thickness and Incisal Relationship on Biomechanical Performance of Narrow Diameter Implant Placed in Atrophic Anterior Maxilla: Finite Element Analysis
Get PDFAtrophic anterior maxilla edentulous space could pose a significant challenge to successful osseointegrated implant due to inadequate labio-palatal dimensions. The load transferring to surrounding bone is a key factor for the long-term success of implant treatment. Thus, the aim of this study was to evaluate the influence of bone quality change in age-related bone mechanical property (AMP), cortical bone thickness (CBT) and incisal relationship (ICR) on the biomechanical performance of narrow diameter implant placed in atrophic anterior maxilla via finite element method. Three-dimensional models of a narrow diameter implant and an anterior maxillary bone were constructed. Eighteen different clinical situations including two CBTs [thin (0.5 mm) and thick (1.0 mm)], three AMPs [young, middle and old ages] under three ICRs [a low overbite (LO), a mean overbite (MO), a high overbite (HO)] were studied under the loading of 50.1 N. From the results, it is crucial to consider the critical situations of narrow diameter implant placed in atrophic anterior maxilla where the combination of the thin CBT, old age-AMP and HO-ICR clinical situation which induce surrounding bone resorption and implant damage
Low Cost Solar Power System with Open Loop Tracking for Rural and Developing Areas
Get PDFSolar energy continues to enjoy interest and is becoming one of the important elements in the world's future energy consumption and economic growth. One way to increase the use of solar power systems, particularly in rural and developing areas, is through the employment of low cost, power efficient systems. This paper presents a possible low cost solution where an open loop tracking system is implemented with a small size 50 Watt (W) Monocrystalline Photovoltaic (PV) panel. The system’s performance is monitored at different tropical weather conditions. In addition, the output power is measured and compared between a static solar panel and one that includes the proposed tracking system. The measurements over the course of 6 months in Pathum Thani, north of Bangkok (Thailand), showed that the output energy with tracking system, on average, is 18 % higher than with static panel (45 Watt hour (Wh) compared to 36.5 Wh). The overall generated energy in 7 hours in a day are 315 Watt hour per day (Wh/day) and 255 Wh/day for tracking and static panel respectively. On rainy days, the average measured output power is reduced to 25% of that on the sunny days. The experimental results have been compared to those of a deduced first order simulation output and this over a period of five months. The experiment shows similarity in the trend between experimental and simulation results. Finally, multiple sensors were added to the solar unit, and measured environmental data sent obtained by these sensors were over a low bandwidth communication link, and recorded/processed on a remote server