Bulletin of Electrical Engineering and Informatics
Not a member yet
2885 research outputs found
Sort by
Internet of things-drone trajectory planning model with edge computing based on long range payload in rural areas
Full text linkThe integration of internet of things (IoT) with unmanned aerial vehicle (UAV) or drone, for precision agriculture (PA) in rural tea plantations is required to ensure optimal outcomes. However, rural settings presents exceptional challenges for data transmission, particularly in maintaining effective communication between drone and ground control stations (GCS). Therefore, this research aimed to develop a payload metadata identification model using long range (LoRa) technology, known for robust IoT capabilities of the model. LoRa was used to transmit drone data packets to GCS, including image data computations and onboard sensor information. Additionally, the research proposed IoT-drone trajectory planning model, specifically designed for PA in rural tea plantations. This model incorporated LoRa technology for data transmission, leveraging the effectiveness of the model in remote areas. Edge computing was also integrated into model to classify the suitability of tea plantation picking areas based on image captured with drone. An important component of the research was trajectory planning system, which optimized drone flight paths by considering location data, throughput data, battery energy consumption, and the computation of suitable picking locations. Finally, experimental results showed the effectiveness of the proposed model in identifying payload metadata, monitoring drone trajectory, and optimizing picking location paths in rural tea plantations
Rainfall forecasting by utilizing adaptive neuro-fuzzy inference system in Aceh Besar District
Full text linkForecasting is a common thing to capture events in future based on previous information. However, some classical time-series methods, including moving average (MA), autoregressive integrated moving average (ARIMA), seasonal autoregressive integrated moving average (SARIMA), and simple exponential smoothing (SES), have limitations in predicting nonlinear time-series data. Therefore, this paper aims to utilize the adaptive neuro-fuzzy inference system (ANFIS) model, a combination of the fuzzy inference system (FIS) and neural network architecture to forecast a nonlinear rainfall problem. This model can capture the non-linear data, adaptation capability, and speedy learning capacity. We used the data consisting of temperature (ºC), humidity (%), and wind speed (km/hour) as input variables and rainfall (millimeter) as an output variable at two stations and one rain post in Aceh Besar District, from January 2009 to December 2019. The results demonstrated that ANFIS with generalized Bell (gBell) membership function on epoch 10 can successfully conduct rainfall forecasting in Aceh Besar District with the best-predicted value. The mean absolute percentage error (MAPE) of the prediction at the Meteorology, Climatology, and Geophysics Agency (MCGA) Station or Badan Meteorologi, Klimatologi dan Geofisika (BMKG) Indrapuri is 6.73% for 80% of the training dataset and 20% of the testing dataset
Comparative analysis of word embedding features to improve the performance of deep learning models on social media data
Full text linkIn this study, we apply various deep learning methods incorporating word embedding features to evaluate their impact on improving classification performance in sentiment analysis. The methods employed include conditional random field (CRF), bidirectional long short term memory (BLSTM), and convolutional neural network (CNN). Our experiments utilize social media data from restaurant review. By testing different iterations of these deep learning techniques with various word embedding features, we found that the BLSTM algorithm achieved the highest accuracy of 80.00% before integrating word embedding features. After incorporating word embeddings, the BLSTM with the word2vec feature achieved an accuracy of 87.00%. Notably, the CNN showed a significant improvement with the FastText feature. Considering all evaluation metrics—accuracy, precision, recall, and F1-score—the BLSTM algorithm consistently demonstrated the best performance across different word embeddings
Solar panel installation feasibility analysis based on techno-economic of PVSyst at Universitas Multimedia Nusantara
Full text linkUniversitas Multimedia Nusantara (UMN) integrates green building principles to enhance environmental sustainability by reducing energy waste and utilizing renewable energy sources. This study conducts a feasibility analysis of installing solar panels in a green open space near building D to supply up to 20% of its electricity needs. PVSyst simulations evaluated different panel orientations (north, south, and east). The results indicated that the installation is currently unfeasible, with a net present value (NPV) of -134,346,450.22 IDR and an internal rate of return (IRR) of -4.64%. The challenges included shading from surrounding buildings, heat buildup, and limited installation space. To improve viability, future installations should focus on sites with minimal shading and explore advanced technologies to enhance efficiency. Additionally, optimizing panel orientation and investigating alternative renewable energy sources suited to UMN’s conditions are crucial. These measures can enhance the effectiveness of solar installations and contribute to overall energy sustainability on campus
Factorized cross entropy integrated hyperspectral CNN (HSCNet-FACE) for hyperspectral image classification
Full text linkThe use of hyperspectral image classification algorithms has garnered increasing interest from the scientific community in recent years, especially in the field of geosciences for pattern recognition applications. In order to extract full spectral-spatial characteristics, this study presents feature extraction with hyperspectral CNN (HSCNet), a unique hierarchical neural network architecture. HSCNet can handle computational complexity issues and capture extensive spectral-spatial information with ease. We use factorized cross entropy (FACE) to address the common problem of class imbalance in both experimental and real-world hyperspectral datasets in order to construct an accurate land cover classification system. FACE makes it easier to reconstruct the loss function, which helps to effectively accomplish the goals that have been expressed. We provide a new framework for hyperspectral image (HSI) classification called FACE, which combines components from HSCNet and FACE. Next, we carry out in-depth studies using two different remote sensing datasets: Botswana (BS) and Indian Pines (IP). We compare the effectiveness of different backbone networks in terms of categorization and compare its classification performance under various loss functions. Comparing our suggested classification system against the state-of-the-art end-to-end deep-learningbased techniques, we find encouraging result
A lightweight convolutional neural network for rice leaf disease detection integrated in an Android application
Full text linkMore than two-thirds of the world's population rely on rice or wheat as staple foods, which are grown in various Asian countries. Diseases affecting rice leaves can disrupt growth, reduce yields, and cause famine in some areas. Therefore, a quick and accurate recognition method is necessary to minimize losses. This article focuses on eight types of rice leaf diseases using data consisting of approximately 110 images for each disease type, with enhanced image quality to achieve better results. The study applies a convolutional neural network (CNN) model integrated into an Android mobile application, achieving a training accuracy of 86.56% and a validation accuracy of 93.75%. Comparative experiments demonstrate that the model can be effectively implemented in mobile applications for accurately detecting rice leaf diseases, providing a reliable solution for field detection. This method not only helps farmers identify diseases more quickly but also has the potential to reduce crop losses caused by leaf diseases
Multi-source assisted mixed simultaneous wireless information and power transfer for energy efficient routing in IoT networks
Full text linkRecently, simultaneous wireless information and power transfer (SWIPT) emerged as the best solution for resource-constrained internet of things (IoT) networks. SWIPT ensures the provision of parallel information and power transfer in the network. Under the SWIPT model, many researchers use two well-established protocols: time switching (TS) and power splitting (PS). TS is better than PS when the signal is weak but inserts an extra delay because energy harvesting (EH) and information decoding (ID) happened two different times. However, PS protocol performs poorly in hard situations with low signal strength even if it conducts EH and ID simultaneously. Hence, this paper proposed a new model called mixed-SWIPT (MSWIPT) which combines TS and PS protocols in an intuitive manner. Further, this work proposes a multi-source EH mechanism in which the receiving node harvests energy from multiple sources which is different from single source, i.e., desired node’s radio frequency (RF) signal. The multiple sources include non-participated Neighbor Node’s RF signal, sink node and co-channel interference and noise. Under the routing, the node selection is formulated as maximum link capacity problems and solved through several constraints. Extensive simulations on proposed model prove the superiority in terms of EH and energy efficiency from the state-of-the-art methods
QEMF for spatial domain pre-processing in iris biometrics: advancing accuracy and efficiency in recognition systems
Full text linkThis article presents a Quantum-Enhanced Median Filtering (QEMF) method for spatial domain pre-processing in iris biometrics, designed to improve image denoising and recognition accuracy. Traditional median filtering often struggles with high noise density, leading to inconsistencies in the denoised image. Our approach enhances the median filtering process by integrating quantum-inspired principles with statistical measures, combining median and average values of neighboring pixels. This hybrid strategy preserves the structural integrity of the original image while effectively reducing noise. Additionally, a quantum-based thresholding step is introduced in the final stage to minimize ambiguities and further enhance image quality. The proposed method is evaluated using approximately one hundred standard iris images from the Chinese University of Hong Kong (CUHK) dataset, considering four types of noise: Impulse, Poisson, Gaussian, and Speckle. Comparative analysis with conventional filters, including Median and Wiener filters, demonstrates that the QEMF method achieves 99.36% similarity to the original images, surpassing Median and Wiener filters by 1.32% and 0.34%, respectively. These results highlight the potential of quantum-enhanced filtering for improved denoising performance and increased efficiency in iris recognition systems
Design and performance evaluation of a high-efficiency circular microstrip patch antenna for RFID applications at 900 MHz
Full text linkThis study presents a high-efficiency circular microstrip patch antenna designed for radio frequency identification (RFID) applications simulation results illustrate the performance of a circular microstrip patch antenna operating at 900 MHz. Microstrip antennas are renowned for their ability to meet the requirements of compact, lightweight designs, ensuring compatibility, and ease of integration. This research focuses on the development of a circular microstrip antenna, formed as a circular patch on a 0.035 mm thick FR-4 substrate. The design was realized using a substrate with a relative permittivity (εr) of 4.3, a loss tangent (tan δ) of 0.021 and a substrate height (h) of 1.6 mm. The antenna dimensions are small, measuring 58×45 mm, with a circular patch radius of 17 mm. The antenna operates over a frequency range from 0.5 GHz to 2 GHz. Key performance parameters include a return loss of -49.8 dB, a wide bandwidth of 150 MHz, a voltage standing wave ratio (VSWR) of 1.009, a gain of 2.161 dB, and a directivity of 2.200 dBi. Antenna design and simulation were carried out using computer simulation technology (CST) Studio Suite Software, specifically adapted to RFID applications
Chili leaf segmentation using meta-learning for improved model accuracy
Full text linkRecognizing chili plant varieties through chili leaf image samples automatically at low costs represents an intriguing area of study. While maintaining and protecting the quality of chili plants is a priority, classifying leaf images captured randomly requires considerable effort. The quality of the captured leaf images significantly impacts the development of the model. This study applies a meta-learning approach to chili leaf image data, creating a dataset and classifying leaf images captured using mobile devices with varying camera specifications. The images were organized into 14 experimental groups to assess accuracy. The approach included 2-way and 3-way classification tasks, with 3-shot, 5-shot, and 10-shot learning scenarios, to analyze the influence of various chili leaf image factors and optimize the classification and segmentation model's accuracy. The findings demonstrate that a minimum of 10 shots from the meta-test dataset is sufficient to achieve an accuracy of 84.87% using 2-way classification meta-learning combined with the mix-up augmentation technique