University of Technology Malaysia
Universiti Teknologi Malaysia Institutional RepositoryNot a member yet
70456 research outputs found
Sort by
Machine learning classification of frequency-hopping spread spectrum signals in a multi-signal environment
No full textFrequency-hopping spread spectrum (FHSS) spreads the signal over a wide bandwidth, where the carrier frequencies change rapidly according to a pseudorandom number making signal classification difficult. Classification becomes more complex with the presence of additive white Gaussian noise (AWGN) and interference due to background signals. In this research, a hybrid convolutional neural network (HCNN) system with the fusion of handcrafted and deep features is proposed to classify FHSS signals in the presence of AWGN and the background signal. The CNN is used as a deep feature extractor by transforming the intermediate frequency (IF) signal to the time-frequency representation (TFR) and used as a two-dimensional (2D) input image, whereas the handcrafted features of the FHSS signal such as hop frequency and hop duration are estimated from the TFR. A proper network structure of the three-layer fully connected network (TLFCN) is determined and used as a classifier. The TLFCN is a machine learning algorithm that requires training with a proper dataset to classify the various types of FHSS signals. Ideally, the dataset size must be sufficiently large as well as balanced to optimize the classification performance. A pseudorandom sequence of hopping frequencies observed from an FHSS signal represents one observation of all the possible hopping sequences of the signal. Therefore, an observation calculating technique is developed that can derive the total number of possible hopping sequences of an FHSS signal by using the frequencies to determine the observations in the dataset. The majority of the machine learning algorithms assume that the training set is evenly distributed among classes. However, in many real-world applications, the number of observations among classes is often imbalanced, which reduces the classification performance of the algorithm. The number of observations of an FHSS signal depends on the number of hop frequencies. Therefore, a given set of FHSS signals with a varying number of hop frequencies among the FHSS signals results in an uneven number of observations, thereby building an imbalanced dataset. Thus, resampling and data augmentation methods such as synthetic minority oversampling technique (SMOTE) and random erasing (RE) are performed to balance the dataset for the increased learning and decision-making capacity of a machine learning algorithm. Monte Carlo simulation is performed to verify the classification performance of the linear discriminant (LD), TLFCN, CNN, and HCNN for various signal-to-noise ratio (SNR) levels. Based on the SNR range at 90% probability of correct classification (PCC) in the presence of AWGN and the background signal, the LD performed worst from 1 to 15 dB among all the methods, whereas the HCNN performed best from -1.58 to -0.66 dB. Moreover, the HCNN with the balanced dataset performed better by 0.14 to 1.06 dB of SNR than with the imbalanced dataset. Therefore, the HCNN system improved the classification performance and performed better than conventional machine learning-based algorithms
Delay model of tumor-immune system interactions with hyperthermia treatment
Full text linkThe interaction of the tumor-immune system was initially based on the immunosurveillance hypothesis that immune cells can identify and kill tumor cells, leading to the use of a prey-predatory model for the description of tumor-immune cell interactions. However, the current biomedical findings reveal a pathway to immunoediting, which hypothesizes the ability of tumors to inhibit, seal, and counteract effector cells. Contrary to the discovery of non-oscillating dynamic biomedicine in solid tumors, existing models show oscillating solutions. Thus, the formulation of an immunoediting model that corresponds to the interaction of the tumor-immune system is sacrosanct in the search for effective malignant tumor treatment. The research suggests an immunoediting delay model of tumor-immune system interactions that combine tumor-immune cytokines derived from tumors to counteract effector cells. Qualitative analysis of this model gives an idea of the conditions for the stability of non-aggressive (benign) tumors and the instability of aggressive (malignant) tumors. The numerical results for these two conditions do not indicate an oscillating solution. Although the elimination of tumors is seen in the case of non-aggressive tumors, the suppression of effector cells and uncontrolled growth of tumors characterize the results for aggressive tumors. To find the best treatment, a sensitivity analysis is performed to ensure the role of the model parameters in the development of the tumor. The analysis reveals the best treatment options to kill tumor cells and strengthen the performance of immune cells. The sensitivity analysis results inform the merger of hyperthermia treatments in the proposed model to investigate the effects of thermal induction on immune cell performance and tumor regression. Discrete-time delays were used to investigate whether hyperthermia treatment was safe for patients who had received other treatments, but no cure occurred. The global stability of hyperthermia treatment is obtained using the Lyapunov function. Furthermore, an optimal heat control strategy for treating malignant tumor hyperthermia is obtained to minimize the effect of heat on normal cells while ensuring the elimination of malignant tumors. This research establishes a unique thermal optimal solution that improves the performance of the effector cell without difficulty
Multimodal convolutional neural networks for sperm motility and concentration predictions
Full text linkManual semen analysis is a conventional method to assess male infertility which includes laboratory technicians examining on parameters such as sperm motility and concentration. Manual evaluation is prone to human errors that causes precision and accuracy issues. The purpose of this research study is to adopt computer vision deep learning techniques and multimodal learning approach in sperm parameters prediction using video-based and image-based input. Convolutional neural network (CNN) has benefited technology industry in recent years, and it has been widely applied in computer vision research tasks as well. Most of the well-established model were designed and pretrained for image-based input, whereas temporal information of video-based input might not be extracted properly using these architectures. Three-dimensional CNN (3DCNN) would be an alternative as it was designed to extract motion and temporal features, which are vital for sperm motility prediction. For sperm concentration, since twodimensional CNN (2DCNN) is efficient in recognizing and extracting spatial features, Residual Network (ResNet) could be adopted for sperm concentration prediction with minimal modification on the original architecture. On the other hand, multimodal learning approach is a technique to aggregate learnt features from different deep learning architecture that adopted other forms of modalities, and provide deep learning model better insights on their tasks. Hence, multimodal learning has been introduced in this research study, where the finalized deep learning architecture received both image-based (frames extracted from video samples) and video-based (stacked frames pre-processed from video samples) input that could provide well-extracted spatial and temporal features for sperm parameters prediction. In this research study, VISEM dataset has been used because it is an open-source dataset which contains 85 sperm videos and biological analysis data from different patients. The video samples went through pre-processing stage to obtain the suitable modalities for training and validation. The developed system has been proven to be capable of improving performance which was as proposed, after the results had been compared to other similar research works. Average mean absolute error (MAE) for sperm motility was observed with high accuracy up to 8.05, and competent performance for sperm concentration with Pearson’s correlation coefficient (RP) of 0.853
Sensor fusion with Kalman filter and support vector machine for fault detection in automated guided vehicle
Full text linkIndustries are moving towards automation and the usage of machines such as Automated Guided Vehicle (AGV) is increasing. Thus, demands for the reliability of AGVs are increasing as they have various complex tasks to carry out. Unfortunately, AGVs are still susceptible to faults and breakdown. Therefore, fault detection is important to provide means of self-diagnosis on AGV. However, fault detections are generally threshold based which are unsatisfying in terms of accuracy and are prone to false triggering. Extended Kalman Filter (EKF) has limitations in handling nonlinear models while Unscented Kalman Filter (UKF) seems promising. Support Vector Machine (SVM) was used as a fault detection method. Thus, this research proposes a sensor fusion enhanced with SVM for fault detection on AGV. The first objective of this research is to develop a test AGV. This AGV is a two-wheeled differential driven mobile robot with multiple sensors and able to make various types of movements to emulate an industrial AGV. Next objective is to develop an enhanced sensor fusion method using EKF and UKF for fault detection with SVM on AGV. The last objective is to evaluate the performance of the developed method. Experiments were carried out where the AGV was used as a test bed for sensor fusion and fault detection. The AGV was tested in different experiment setups such as different track layout, different wheel condition, and different castor conditions. Result shows that UKF handles changes and non-linearity better than EKF. The average residual generated during the test for UKF is 0.0083 meter while for EKF is 0.0129 meter. With sensor fusion, deviations in odometry data can be compensated with the usage of a LiDAR sensor as reference. Using UKF parameters to detect fault, the accuracy achieved with SVM is 64.2% compared to 37.9% without SVM. Fault detection accuracy using EKF parameters with SVM is 82.5% while without SVM is 41.0%. As a conclusion, the results show SVM improves fault detection accuracy regardless of using UKF or EKF
Light-emitting diode lighting parametric study of internally illuminated photobioreactor for good growth of nannochloropsis species cultivation
No full textMicroalgae such as Nannochloropsis sp. are single-cell organism, well known as a photosynthetic microorganism and have been identified as one of the most potential renewable feedstock for the third generation biofuels. Light quality and quantity are essential for a good growth of microalgae cultivation. In this research, lighting parametric study using light-emitting diode (LED) was conducted to analyse the effect of different light spectrum onto the growth of Nannochloropsis sp. cultivation. In addition, the effect of light intensity with different optical path length on different working volume culture was also evaluated. In order to validate a good growth of cultivation, culture growth curve was analysed and maximum cell density was recorded. The study was performed in three stages of experimental photobioreactor (PBR) setup which is lab-scale, mock-up and scale-up PBR. In the lab-scale experiment (0.5 working volume), LED with red spectrum (wavelength 660 ) and blue spectrum (wavelength 457 ) were compared to the white fluorescent light with same incident light intensity (100 at short optical path length (20 to 55 ). It was found that LED with combination of red and blue spectrum generated higher maximum cell density by 19% compared to the white fluorescent light, which recorded at 11.2 106 . In the mock-up PBR experiment (20 working volume), light intensity of red and blue LED module with narrow beam angle (55 ) was evaluated within 15 to 120 of optical path length by using variation of current supply (200 to 500 ). As the result, the maximum cell density recorded is 7.1 106 at 355 of light intensity. Additionally, the relationship between light intensity and culture cell density was also established at this stage. Next, the cultivation was performed continuously in the scale-up PBR with bigger working volume (30, 65 and 100 ) at 100 optical path length. It was found that light saturated condition happened at cell density around 7.5 106 to 8.0 106 when the light intensity is at 350 . The maximum cell density can be increased further to 9.3 106 by applying higher light intensity (450 ). As a conclusion, LED with red spectrum promoted the growth in the exponential growth phase, while blue spectrum had a significant role during the linear growth phase especially in the higher cell density culture. Application of high intensity LED light with narrow beam angle was feasible to be used in internally illuminated PBR with longer optical path length. On top of these findings, a vertically stackable LED luminaire design concept was proposed to provide flexibility and to increase efficiency for mass cultivation operation using internally illuminated PBR
Assessment of Blastability Index in massive limestone from Rawang Quarry, Selangor
Full text linkThe demand for construction materials produced by quarry rises in tandem with urbanization. Selangor is one of the highly populated states in Malaysia with residential projects now located very close or even next to quarry. Due to the fact that limestone consists of numerous weak spots in rock masses, it has always been thought that limestone quarry operation is riskier than common granite quarry operations. Geologically, limestone formation in Rawang sits uncomfortably on top of the older metamorphic rocks with its own unique joints system. The goal of the study was to identify the rock mass properties in massive limestone profile from a quarry in Rawang, Selangor and its relation to blast design as well as effects on the surrounding environment due to blasting. For a systematic study, the quarry face was divided into four (4) sections i.e., section A, section B, section C, and section D. The site mapping showed significant findings where section A is considered high potential of having excessive flyrock as it has the most joint number (J), joint plane spacing (JPS) and joint aperture (JA) with 31, 559.8 mm and 28.5 mm, respectively. When blasthole intersected with many joints, explosive energy escape through joints causing sudden drop in blasthole pressure and open joints extend up to the face thus creating high possibility of flyrock during blasting. The degree of difficulty to fragment rock in terms of Blastability Index (BI) was also calculated based on the geological mapping data. The results show that BI ranged from 49.18 to 59.26 percent throughout all study sections indicating that the rock mass at the quarry was easy to be blasted as per Blastibility Quality System (BQS). The calculated BI was also justified the suitability of blast design used during blasting at the quarry. The new site constants i.e., K and β for the study area were also calculated with USBM predictor at 40 and 1.0, while Langefors- Kihlstrom (LK) predictor at 6.8 and 1.07, respectively. Although at maximum charge per delay (Wmax) the blasting was being carried out safely with very minimal effects to the surrounding areas. Finally, correlations between all earlier findings such as BI, blast design and environment effects i.e., peak particle velocity (PPV) measured and predicted were justified the significant relation of rock mass properties and ground vibration effect due to blasting operation at the quarry
Synthesis, characterization and fire retardancy of titania-based materials coated on wood
Full text linkThe number of fire cases in Malaysia has recently been increasing year after year, necessitating fire prevention measures. One prevention method is introducing fire retardant (FR) coating materials. FR coating material is a coating layer that works to prevent and reduce the probability of the material being flammable. Titania (TiO2) nanoparticles (NPs) and silica/titania (SiO2/TiO2) nanocomposites (NCs) worked as the FRs coating materials in this project, while rubberwood functioned as the flammable substrate material. The fire performance behaviour was then investigated using cone calorimeter, thermogravimetric analyzer (TGA), and flammability tests. According to the findings, FR coating of TiO2 NPs and SiO2/TiO2 (ratio of 0.1:1) NCs were able to increase the decomposition temperature (OD) of rubberwood by 41.32°C and 37.59°C, spontaneous ignition (SI) by 45.95°C and 32.6°C, and delayed ignition time (IT) by 79 s and 114 s. The results also proved the reduction in the intensity of fire (FI), heat release rate (HRR) by 43.32 kW/m2 and 45.87 kW/m2, rate of combustion (ROC) by 0.144 mm/s and 0.142 mm/s, mass loss rate (MLR) by 2.7 g/s and 2.9 g/s, and combustion efficiency (EHC) by 4.89 mJ/kg and 4.95 mJ/kg, respectively. The fuzzy logic system determined from these parameters that the HRR parameter should be considered as the key parameter that needs to be decreased in order to improve the FR performance of the TiO2-based materials. The physicochemical properties of TiO2-based materials coated on wood were then analyzed by various instruments and methods. Field emission scanning electron microscopy (FESEM) showed that the TiO2-based materials coated on the wood are spherical in shape, in the nano-range (25 to 40 nm), and agglomerated on the surface of the rubberwood. Meanwhile, energy dispersive X-ray (EDX) confirmed that the presence of titanium (Ti) and silica (Si) as the primary elements. The principal functional groups of TiO2 and SiO2 were also visible by fourier transform infrared (FTIR) spectroscopy. The X-ray diffraction patterns (XRD) demonstrated that the TiO2 in both TiO2 and SiO2/TiO2 samples are present in the anatase phase with a lower crystallinity, corresponding to the bandgap energies (3.2 eV and 3.4 eV) determined using diffuse reflectance ultraviolet-visible spectroscopy (UV-Vis). Through the peel adhesion test, it was proven that the application of 3-aminopropyltrimethoxysilane (APTMS) as a surface modifying agent allowed strong adhesion between the TiO2-based materials and wood and was uniformly coated. The wettability test showed that the surface of the rubberwood changed from superhydrophilic to hydrophilic. In conclusion, this study has demonstrated that TiO2 and SiO2/TiO2 (0.1:1) are the best coating materials on wood that can successfully operate as fire retardant materials, displaying the potential to decrease the performance of wood burning
Optimisation of Arthrospira Platensis harvesting using edible fungal bioflocculant
No full textArthrospira sp. is considered a sustainable and completely natural microalgae-based food supplement to solve nutritional diseases, specifically malnutrition. However, both cultivation and harvesting methods for this microalgae takes up to 40% of energy consumption. Therefore, this study is aimed at maximizing Arthospira platensis biomass productivity under outdoor cultivation as well as propose safe and efficient harvesting method using edible fungi. Results from data obtained by comparing three types of photobioreactor (PBR) configurations conducted under indoor conditions demonstrated that macrobubble column (MA-CP) showed the highest dry cell weight yield compared to microbubble column (MI-CP) and airlift loop column (ALCP), with 0.536 ± 0.044 g/L, 0.477 ± 0.034 g/L and 0.274 ± 0.014 g/L, respectively. Thus, based on this result, MA-CP was carried out during the outdoor cultivation studies. Covered MA-CP showed a comparable but steady growth compared to non-covered MA-CP due to limited exposure of the microalgae to solar radiation. Whereas outdoor MA-CP PBRs resulted in significantly higher growth compared to indoor MA-CP due to the influence of temperature and light intensity. The result suggested that by taking advantage of Malaysia's weather conditions, integration of solar panel systems for outdoor cultivation of Arthrospira sp. in covered MA-CP is a viable and sustainable option. Meanwhile, a promising harvesting technique via bioflocculation is recommended as an alternative to conventional flocculation because of its simplicity and efficiency. In this study, Rhizopus microsporus was locally isolated and demonstrated the highest harvesting efficiency compared to other fungi. One-factor-at-time (OFAT) technique was used for the preliminary screening of different factors including mycelia concentration, pH of mycelia and temperature. The results were then applied in response surface methodology (RSM) modelling for optimization through central composite design (CCD). The harvesting efficiency (HE, %) for bioflocculation of A. platensis using R. microsporus was maximized (65.89 ± 2.795%) when 3.85% mycelial concentration (w/v) with initial pH of 2.5 at 38.8°C were used as the harvesting parameter conditions. Overall, the results showed that the overall process is viable and economical when the outdoor cultivation setup was integrated with solar panels as the system produced 2-fold biomass compared to the indoor cultivation coupled with harvesting microalgae step via locally isolated fungi as bioflocculant
An improved corrosion resistance of mild steel in acidic solution using aloe vera extract
No full textCorrosion is a natural process and unfavourable occurrence when pipeline metals and alloys react with their surrounding environment, cause undesirable damage, and create harmful environment. This issue has resulted in substantial economic losses and posed significant safety concerns. In order to overcome these problems, typically, the industry employs the use of green inhibitors corrosion to safeguard metal surfaces from corrosive agents. However, commercial inhibitors pose a danger and toxic to both the environment and human health. Therefore, as a solution to those problem, green corrosion inhibitor has been found to be both effective and economically feasible, performing similarly to commercial inhibitors. This study aims to examine the efficacy of Aloe Vera extract as a green inhibitor of corrosion in mild steel in a 1M H2SO4 solution. Weight loss, Potentiodynamic polarization, Electrochemical Impedance Spectroscopy (EIS) and Scanning Electron Microscopy (SEM) studies were conducted in a 1M H2SO4 solution towards mild steel to analyse the effect of Aloe Vera as an effective corrosion inhibitor. It has been found that as the concentration of inhibitors increases, inhibitor efficiency also increases. Weight loss measurement has proven that the inhibition efficiency increased and reached to maximum values of 90.81 % after 8 h exposure time for 800 ppm concentration. Potentiodynamic polarization result also shown similar trend with weight loss where the corrosion rate is reduced with inhibitive efficiency with higher concentration of Aloe Vera content in the acidic solution. The result obtained from EIS shows one depressed capacitive loop where the capacitive loops diameter was bigger in presence of higher Aloe Vera content which suggested bigger imaginary film that protecting the surface. SEM analysis revealed a significant decrease in the roughness of mild steel surface after the addition of an inhibitor, which support the inhibitor efficiency of protecting the mild steel surface from the corrosive medium
Preparation of solid biofuel derived from palm kernel shell waste via pyrilysis and hydrogenation process
Full text linkThe need for energy is always expanding along with industrialization and population increase, and the availability of energy supplies cannot keep up with the mounting consumption. In 2000, one estimate predicted that about 20 million metric tons of CO2 will be released into the atmosphere yearly. While this pattern continues, extreme natural catastrophes, such as excessive rainfall and the ensuing floods, droughts, and local imbalance, are to be expected. The world's oil reserves are expected to run out by 2050. Considering the mentioned aspects there has been an elevating demand for renewable resources of energies. One of such mentionable sources of energy is palm oil leftovers or palm kernel shells (PKS). Low moisture content, compact and having a high calorific content are some of the mentionable factors contributing to the popularity of PKS. Considering the mentioned aspects, in this study, the potential of palm kernel shell waste as a solid biofuel by using pyrolysis and hydrogenation process has been evaluated along with analysing the impact of temperature in achieving a higher calorific value. In order to evaluate the heating values of the samples, Bomb calorimeter has been used and TGA analysis has been performed to evaluate the impact of temperature and associated changes in the weight of the sample. For analysing the compound group present in the samples, FTIR analysis has been conducted. The highest heating value of 25.96 MJ/kg was obtained for sample prepared at 300oC hydrogenation temperature at 1 hour with the use Ru/Mn/ce-65 as catalyst. Significant C-H stretching and vibration was observed at 2850 – 3000 cm-1 and 1120-1250 cm-1 wavelength which indicate increasing C-H adsorption at carbon surface that become a major contribution to higher heating value. In addition to that, temperature has been observed to have major impact on the performance of the samples in attaining higher calorific value as compared to reaction time. Therefore, PKS proved to have a great potential as green waste to be used as a high-quality solid biofuel for combustion purposes