Taiwan Association of Engineering and Technology Innovation: E-Journals
Not a member yet
887 research outputs found
Sort by
Comparative Study of Relay Coordination in a Microgrid with the Determination of Common Optimal Settings Based on Different Objective Functions
Get PDFThis study aims to analyze the optimal settings of directional overcurrent relays (DOCRs) for the protection of an alternating current (AC) microgrid in both islanded and grid-connected operation modes. In this context, two different types of objective functions are used for comparing the total operating time of all primary DOCRs. The optimal settings obtained in either mode of the microgrid are different due to the variable magnitude of fault currents. The proposed protection coordination scheme is formulated as a mixed-integer non-linear programming problem, and the settings are obtained using various optimization techniques such as firefly algorithm, simulated annealing algorithm, and genetic algorithm. The results show that the settings obtained in common operation modes are robust as no miscoordination of relays occurs in any of the operation modes
Intelligent Correction and Monitoring of Ship Propulsion Rotary Device Vibration
Get PDFField inspection is a traditional way to detect the problem of shaft imbalance or abnormal vibration in a ship propulsion system; however, the ship cannot execute any tasks or activities during calibration. This study develops a human-machine monitoring interface (HMMI) to estimate vibration abnormalities and implement an intelligent active balance correction to the propulsion system online. In this study, Arduino IDE, InduSoft, and LabVIEW are used to create a function monitored by HMMI. By comparing the abnormal vibration amplification of the moment of inertia, HMMI calculates the correct mass to reduce the vibration. The experimental results show that, after HMMI carries out continuous active balance correction online, the correction rate achieves 105.37%. This indicates that HMMI can calculate the amount of imbalance and phase angles and drive a counterweight to the correct balance position while the device is still operating
Prediction of Wind Turbine Airfoil Performance Using Artificial Neural Network and CFD Approaches
Get PDFTo achieve the highest energy level from a wind turbine, the prediction of its performance is essential. This study investigates the aerodynamic performance of different airfoils, which are frequently used in wind farms. The computational fluid dynamics method based on the finite-volume approach is utilized, and a steady-state flow with the transition regime is considered in this study. A developed artificial neural network is used to reduce the computational time. The results indicates that the trained algorithm could accurately predict the airfoil efficiency with less than 2% error on the training set and fewer than 3% error on the test set. The results agree with the experimental results; this analysis accurately predicts wind turbine performance by selecting the blade’s airfoil. This study provides a reference for a broader range of using these airfoils in wind farms
A Solar Energy Harvester for a Wireless Sensor System toward Environmental Monitoring
Get PDFHarmful environments can cause severe health problems to individuals. Thus, this study proposes a solar-powered wireless sensor system to monitor the physical parameters of an ambient environment in real-time. This system is developed based on two sensors and a NodeMCU board that includes a microcontroller with a Wi-Fi chip. This system is built to measure the ambient temperature, relative humidity, atmospheric pressure, and ultraviolet (UV) index. The power supply of the system is a solar energy harvester, which consists of a solar cell, a DC-DC converter, and a rechargeable battery. This harvester is practically tested outdoors under direct sunlight. The proposed system experimentally consumes an average power of 40 mW over one hour, and the lifetime of this system is 123 hours in the active-sleep mode. The results demonstrate that the system can sustainably operate for monitoring the environmental data
An Integrated Approach towards Efficient Image Classification Using Deep CNN with Transfer Learning and PCA
Get PDFIn image processing, developing efficient, automated, and accurate techniques to classify images with varying intensity level, resolution, aspect ratio, orientation, contrast, sharpness, etc. is a challenging task. This study presents an integrated approach for image classification by employing transfer learning for feature selection and using principal component analysis (PCA) for feature reduction. The PCA algorithm is employed for reducing the dimensionality of the features extracted by the VGG16 model to obtain a handful of features for speeding up image reorganization. For multilayer perceptron classifiers, support vector machine (SVM) and random forest (RF) algorithms are used. The performance of the proposed approach is compared with other classifiers. The experimental results establish the supremacy of the VGG16-PCA-Multilayer perceptron model integrated approach and achieve a reorganization accuracy of 91.145%, 95.0%, 92.33%, and 98.59% on Fashion-MNIST dataset, ORL dataset of faces, corn leaf disease dataset, and rice leaf disease datasets, respectively
Experimental and Analytical Study of Silica Particles on Self-Healing Concrete
Get PDFThis study aims to investigate the properties of green concrete made with ground granulated blast-furnace slag (GGBS), Robo sand (RS), and coconut shell (CS). GGBS is the mineral admixture used to replace cement. Nano-silica particles (NSPs) and CS are used as coarse aggregates, and RS is the fine aggregate used to replace river sand. The workability, mechanical properties, and durability properties of green concrete are investigated and compared with those of conventional concrete (CC). Test results show that the cement replaced with 30% GGBS and 3% NSPs exhibits superior strength. The compressive and splitting tensile strengths are increased by 24.03% and 42.32% after 28 days of curing, respectively. The workability is improved by 12.22% (slump) and 13.25% (compaction factor) after 28 days of curing. The sorptivity of HM3 (3.26%) is lower than that of CC due to the uniform distribution between particles. Microstructure evolution is carried out to identify concrete mix behavior
Lightweight Compressive Sensing for Joint Compression and Encryption of Sensor Data
Get PDFThe security and energy efficiency of resource-constrained distributed sensors are the major concerns in the Internet of Things (IoT) network. A novel lightweight compressive sensing (CS) method is proposed in this study for simultaneous compression and encryption of sensor data in IoT scenarios. The proposed method reduces the storage space and transmission cost and increases the IoT security, with joint compression and encryption of data by image sensors. In this proposed method, the cryptographic advantage of CS with a structurally random matrix (SRM) is considered. Block compressive sensing (BCS) with an SRM-based measurement matrix is performed to generate the compressed and primary encrypted data. To enhance security, a stream cipher-based pseudo-error vector is added to corrupt the compressed data, preventing the leakage of statistical information. The experimental results and comparative analyses show that the proposed scheme outperforms the conventional and state-of-art schemes in terms of reconstruction performance and encryption efficiency
Deep Learning for Image-Based Plant Growth Monitoring: A Review
Get PDFDeep learning (DL) approaches have received extensive attention in plant growth monitoring due to their ground-breaking performance in image classification; however, the approaches have yet to be fully explored. This review article, therefore, aims to provide a comprehensive overview of the work and the DL developments accomplished over the years. This work includes a brief introduction on plant growth monitoring and the image-based techniques used for phenotyping. The bottleneck in image analysis is discussed and the need of DL methods in plant growth monitoring is highlighted. A number of research works focused on DL based plant growth monitoring-related applications published since 2017 have been identified and included in this work for review. The results show that the advancement in DL approaches has driven plant growth monitoring towards more complicated schemes, from simple growth stages identification towards temporal growth information extraction. The challenges, such as resource-demanding data annotation, data-hungriness for training, and extraction of both spatial and temporal features simultaneously for accurate plant growth prediction, however, remain unsolved
An Experimental Study on the Mechanical Properties of Low-Aluminum and Rich-Iron-Calcium Fly Ash-Based Geopolymer Concrete
Get PDFLimited studies have been conducted on low-aluminum and rich-iron-calcium fly ash (LARICFA)-based geopolymer concrete with increased strength. This study aims to investigate the mechanical characteristics of LARICFA-based geopolymer concrete, including its compressive strength, split tensile strength, and ultimate moment. The steps of this study include material preparation and testing, concrete mix design and casting, specimen curing and testing, and the analysis of testing results. Furthermore, the specimen tests consist of the bending, compressive, and split tensile strength tests. The results show that the average compressive strength and the ultimate moment of the geopolymer concrete are 38.20 MPa and 22.90 kN·m, respectively, while the average ratio between the split tensile and compressive strengths is around 0.09. Therefore, the fly ash-based geopolymer concrete can be used in structural components
SOM-FTS: A Hybrid Model for Software Reliability Prediction and MCDM-Based Evaluation
Get PDFThe objective of this study is to propose a hybrid model based on self-organized maps (SOM) and fuzzy time series (FTS) for predicting the reliability of software systems. The proposed SOM-FTS model is compared with eleven traditional machine learning-based models. The problem of selecting a suitable software reliability prediction model is represented as a multi-criteria decision-making (MCDM) problem. Twelve software reliability prediction models, including the proposed SOM-FTS model, are evaluated using three MCDM methods, four performance measures, and three software failure datasets. The results show that the proposed SOM-FTS model is the most suitable model among the twelve software reliability prediction models on the basis of MCDM ranking