Search

Solar energy utilization, in conjunction with battery systems, within stand-alone DC microgrid systems represents a significant trend. In such isolated network configurations, the voltage of the Direct Current (DC) bus experiences fluctuations: it increases when the islanded DC microgrid receives excess energy and decreases during periods of energy scarcity. Ensuring the appropriate operational mode for photovoltaic (PV) panels is crucial for maintaining the DC bus voltage within specified operational limits, guaranteeing high electrical quality, whether the power is supplied directly or converted to AC form. Furthermore, achieving balanced power sharing between the PV and the energy storage system enhances the efficiency of the energy storage workload. This study proposes a transition function to facilitate seamless switching of the PV panel’s operational mode between Maximum Power Point Tracking (MPPT) and voltage-controlled modes. This transition function is applied to the PV side converter, which adjusts the duty cycle value of both MPPT and droop mode outputs based on the microgrid’s energy adequacy condition, as determined by the DC bus voltage readings. Moreover, this study introduces a Duty Cycle Range Divider (DCRD) algorithm to derive the converter’s Piece-wise Affine (PWA) model. Subsequently, a linear quadratic regulator (LQR) controller, designed based on the PWA model, is employed alongside the transition function to enhance the DC microgrid’s dynamic performance. A similar LQR controller is applied to the battery-side converter with a battery State of Charge (SoC)-based droop control to balance the power-sharing. The proposed control strategy stabilizes the DC bus voltage and ensures a seamless response during transitions in the PV system’s operating mode. The efficacy of this strategy is validated through MATLAB Simulink simulations and laboratory-scale experiments.

This study proposes a path planning algorithm that combines an informed rapidly-exploring random tree (RRT*) and particle swarm optimization (PSO) algorithms. The informed RRT* algorithm can build optimal pathways quickly. However, in some cases, the informed RRT* strategy cannot help increase the convergence rate. Meanwhile, the PSO algorithm can be combined with other algorithms to improve algorithm performance. The proposed algorithm will combine the advantages of informed RRT* and PSO algorithms. The proposed algorithm is called the RRT-PSO algorithm. The RRT-PSO algorithm’s performance was compared to the informed RRT*, RRT*, RRT-ACS, and informed RRT*-connect algorithms using eight benchmark scenarios. The test results on benchmark scenarios with known optimal values demonstrate that the RRT-PSO algorithm achieves those optimal values. The statistical analysis shows that the RRT-PSO algorithm has a fast convergence rate compared to other comparison algorithms. The test results also demonstrate the RRT-PSO algorithm’s stability and robustness compared to other comparison algorithms. Our test results also show the RRT-PSO algorithm’s advantages over other PSO-based and traditional path planning algorithms. The results show that the RRT-PSO algorithm is suitable for applications requiring a fast and optimal path planning algorithm, such as self-driving cars and unmanned aerial vehicles (UAV).

This study aims to automatically analyze and extract abnormalities in the lung field due to Coronavirus Disease 2019 (COVID-19). Types of abnormalities that can be detected are Ground Glass Opacity (GGO) and consolidation. The proposed method can also identify the location of the abnormality in the lung field, that is, the central and peripheral lung area. The location and type of these abnormalities affect the severity and confidence level of a patient suffering from COVID-19. The detection results using the proposed method are compared with the results of manual detection by radiologists. From the experimental results, the proposed system can provide an average error of 0.059 for the severity score and 0.069 for the confidence level. This method has been implemented in a web-based application for general users. • A method to detect the appearance of viral pneumonia imaging features, namely Ground Glass Opacity (GGO) and consolidation on the chest Computed Tomography (CT) scan images. • This method can separate the lung field to the right lung and the left lung, and it also can identify the detected imaging feature's location in the central or peripheral of the lung field. • Severity level and confidence level of the patient's suffering are measured.

Autonomous vehicles rely on the accurate detection and recognition of objects in their surroundings, a critical requirement for safe operation, especially in congested traffic with diverse vehicle types. This study presents a novel dataset collected in various road conditions in Indonesia, and it focuses on the detection and classification of visual objects around autonomous vehicles. Object recognition is achieved through the use of YOLOv7-based deep learning, adapted to identify small, faint, and partially concealed objects. Key enhancements include the integration of a deformable layer and the transition from Non-Maximum Suppression (NMS) to Soft Non-Maximum Suppression (softNMS). The dataset comprises eight predefined custom classes commonly encountered in Indonesian traffic. We collected video data recordings of heavy traffic scenarios featuring a wide range of vehicle types as training and testing data. The object detection model is fine-tuned through transfer learning, with multiple learning configurations explored for comparison. Experimental results demonstrate that deep learning models trained with transfer learning outperform those trained from scratch. Specifically, the modified YOLOv7, referred to as YOLOv7-MOD, incorporates a deformable convolution layer for up-sampling, leading to a remarkable performance of 94.68% in Recall, 96.87% in Precision, and 95.76% in F1-score. The modification resulted in an additional performance increase of 1.05% on average compared to the original model. The findings indicate that YOLOv7-MOD enhances the precision of object detection and recognition compared to the original YOLOv7, making it a promising solution for autonomous vehicle perception systems.

Thanks to the complementary nature of millimeter wave radar and camera, deep learning-based radar-camera 3D object detection methods may reliably produce accurate detections even in low-visibility conditions. This makes them preferable to use in autonomous vehicles’ perception systems, especially as the combined cost of both sensors is cheaper than the cost of a lidar. Recent radar-camera methods commonly perform feature-level fusion which often involves projecting the radar points onto the same plane as the image features and fusing the extracted features from both modalities. While performing fusion on the image plane is generally simpler and faster, projecting radar points onto the image plane flattens the depth dimension of the point cloud which might lead to information loss and makes extracting the spatial features of the point cloud harder. We proposed ClusterFusion, an architecture that leverages the local spatial features of the radar point cloud by clustering the point cloud and performing feature extraction directly on the point cloud clusters before projecting the features onto the image plane. ClusterFusion achieved the state-of-the-art performance among all radar-monocular camera methods on the test slice of the nuScenes dataset with 48.7% nuScenes detection score (NDS). We also investigated the performance of different radar feature extraction strategies on point cloud clusters: a handcrafted strategy, a learning-based strategy, and a combination of both, and found that the handcrafted strategy yielded the best performance. The main goal of this work is to explore the use of radar’s local spatial and point-wise features by extracting them directly from radar point cloud clusters for a radar-monocular camera 3D object detection method that performs cross-modal feature fusion on the image plane.

An autonomous underwater vehicle is a vehicle that can move in water, which is also known as an unmanned undersea vehicle. One type is the hybrid underwater glider where the vehicle is designed in such a way that it is able to carry out missions in the water with less power consumption so that it can last a long time in carrying out missions. In this research, a mathematical design is carried out in the form of a nonlinear model with the aim of being able to produce a model in the longitudinal movement of the HUG vehicle which will be tested limited to a simulation using the MATLAB/Simulink program. The parameters used in the model for this longitudinal movement are obtained by the computational fluid dynamics method so that it has been simulated with various movements according to the mission of the vehicle. In the simulation, input is given in the form of variations in the value of the actuator force to be able to carry out movements according to the mission and the simulation is open loop so that the vehicle's response is in the form of position and speed of translation and rotation.

A decision-making system is an essential component of an autonomous tram. The decision-making system uses information on the surrounding area to recognize the traffic situation and determine appropriate actions to maintain safety and passenger comfort. However, guaranteeing reliable performance and safety remains challenging in all driving situations. The development of autonomous trams involves iterative engineering-related tasks. Therefore, proper architecture to facilitate the flexibility of engineering development is required. This paper proposes a modular architecture for a decision-making system for autonomous trams. The decision-making system can serve as a high-level controller for autonomous trams. The architecture consists of risk assessment and decision & planning modules. It also integrates several key functions, such as trajectory prediction, safety assessment, adaptive cruise control (ACC), collision avoidance (CA), and emergency braking system (EBS). In the decision and planning module, a finite-state machine is devised as part of the decision-making system. This module provides a speed reference for low-level speed controllers. In addition, the decision-making system architecture was implemented and its implementation was validated using a Carla simulator. Under a mixed-traffic scenario, simulation results showed that the decision-making system has a high percentage of mission successes. Out of 50 simulations in mixed-traffic scenarios, the tram reached its destination with an 80% success rate in which the success rates of ACC, CA, and EBS executions to avoid collisions were 96.94%, 100%, and 100%, respectively. In addition, the system works well in real-time to recognize the surrounding environment and determine actions. Although beyond the scope of the decision-making system, the simulation results also indicate that an improving performance of the tram’s low-level speed controller may provide more reliable performance of the decision-making system.

This paper presents the implementation of an adaptive control approach to the ball and beam system (BBS). The dynamics of a BBS are non-linear, and in the implementation, the uncertainty of the system's parameters may occur. In this research, the linear state-feedback model reference adaptive control (MRAC) is used to synchronize the states of the BBS with the states of the given reference model. This research investigates the performance of the MRAC method for a linear system that is applied to a non-linear system or BBS. In order to get a faster states convergence response, we define the initial condition of the feedback gains. In addition, the feedback gains are limited to get less oscillation response. The results show the error convergence is improved for the different sets of the sinusoidal reference signal for the MRAC with modified feedback gains. The ball position convergence improvement of MRAC with modified feedback gains for sinusoidal reference with an amplitude of 0.25, 0.5, and 0.75 are 35.1 %, 36 %, and 52.4 %, respectively.

This paper proposes a path planning algorithm using the hybridization of the rapidly-exploring random tree (RRT) and ant colony system (ACS) algorithms. The RRT algorithm can quickly generate paths. However, the resulting path is suboptimal. Meanwhile, the ACS algorithm can generate the optimal path from the suboptimal previous path information. Then, the proposed algorithm will combine the advantages of RRT with the ACS algorithm. Therefore, it can reach the optimal value with a good convergence speed. We call this proposed algorithm the RRT-ACS algorithm. This study developed a new method for hybridizing the RRT and ACS algorithms for path planning problems. This hybridization process is carried out using one of the ACS principles: the pseudorandom proportional rule. The performance of the proposed algorithm with the RRT*, informed RRT*, RRT*-connect, and informed RRT*-connect algorithms is tested with several benchmark cases. The test results from benchmark case tests with known optimal values indicate that the proposed algorithm has succeeded in achieving those optimal values. Furthermore, statistical tests have also been carried out to verify whether there is a significant difference in performance between the RRT-ACS algorithm and the existing algorithms. The test and statistical analysis results show that the RRT-ACS algorithm has good performance and convergence speed. We also discuss the stability, robustness, convergence, and rapidity of the RRT-ACS algorithm. The results indicates that the RRT-ACS algorithm may be used in applications that require fast and optimal path planning algorithms such as robots and autonomous vehicles.

The main goal of this paper was to design and construct a hybrid autonomous underwater glider (HAUG) with a torpedo shape, a size of 230 cm in length and 24 cm in diameter. The control, navigation, and guidance system were executed simultaneously using a Udoo X86 minicomputer as the main server and three BeagleBone Black single-board computers as the clients. The simulations showed a controlled horizontal speed of 0.5 m/s in AUV mode and 0.39 to 0.51 m/s in glide mode with a pitch angle between 14.13∘ and 26.89∘. In addition, the field experiments under limited space showed the proposed HAUG had comparable results with the simulation, with a horizontal speed in AUV mode of 1 m/s and in glide mode of around 0.2 m/s. Moreover, the energy consumption with an assumption of three cycles of gliding motion per hour was 51.63 watts/h, which enabled the HAUG to perform a mission for 44.74 h. The proposed HAUG was designed to hold pressure up to 200 m under water and to perform underwater applications such as search and rescue, mapping, surveillance, monitoring, and maintenance.

Considering the increasing use of security and surveillance systems, moving object tracking systems are an interesting research topic in the field of computer vision. In general, a moving object tracking system consists of two integrated parts, namely the video tracking part that predicts the position of the target in the image plane, and the visual servo part that controls the movement of the camera following the movement of objects in the image plane. For tracking purposes, the camera is used as a visual sensor and applied to a 2-DOF (yaw-pitch) manipulator platform with an eye-in-hand camera configuration. Although its operation is relatively simple, the yaw-pitch camera platform still needs a good control method to improve its performance. In this study, we propose a moving object tracking system on a prototype yaw-pitch platform. A m-synthesis controller was used to control the movement of the visual servo part and keep the target in the center of the image plane. The experimental results showed relatively good results from the proposed system to work in real-time conditions with high tracking accuracy in both indoor and outdoor environments.

Target market classification is aimed at focusing marketing activities on the right targets. Classification of target markets can be done through data mining and by utilizing data from social media, e.g. Twitter. The end result of data mining are learning models that can classify new data. Ensemble methods can improve the accuracy of the models and therefore provide better results. In this study, classification of target markets was conducted on a dataset of 3000 tweets in order to extract features. Classification models were constructed to manipulate the training data using two ensemble methods (bagging and boosting). To investigate the effectiveness of the ensemble methods, this study used the CART (classification and regression tree) algorithm for comparison. Three categories of consumer goods (computers, mobile phones and cameras) and three categories of sentiments (positive, negative and neutral) were classified towards three target-market categories. Machine learning was performed using Weka 3.6.9. The results of the test data showed that the bagging method improved the accuracy of CART with 1.9% (to 85.20%). On the other hand, for sentiment classification, the ensemble methods were not successful in increasing the accuracy of CART. The results of this study may be taken into consideration by companies who approach their customers through social media, especially Twitter.

Wireless Sensor Network (WSN) is small embedded devices deployed in large scale network and has capability to sense, compute, and communicate with others. It combines modern sensor, microelectronic, computer, communication, and distributed processing technology. It takes into account in several aspects of live especially in structural health monitoring system of bridge. Due to environmental circumstance, a bridge should be monitored to make sure that it can perform its service safely. Therefore, it is necessary to develop WSN application to monitor bridge condition and send warning message to control room if the bridge is under abnormal condition. This paper proposes the development of automatic WSN system for measuring the level of bridge structural health based on its dynamic responses. The main contribution of this work is developing WSN system for vibration-based bridge condition assessment in which identifies the bridges fundamental frequency and mode shape. Experimental result shows that the fundamental frequency measured by our proposed system is close to the value analyzed using finite element analysis (FEA) and according to Modal Assurance Criteriation (MAC), our proposed measurement system has correlation with FEA

This paper presents a new formulation for obstacle and collision behavior on a group of humanoid robots that adopts walking behavior of pedestrian crowd. A pedestrian receives position information from the other pedestrians, calculate his movement and then continuing his objective. This capability is defined as socio-dynamic capability of a pedestrian. Pedestrian’s walking behavior in a crowd is an example of a sociodynamics system and known as Social Force Model (SFM). This research is trying to implement the avoidance terms in SFM into robot’s behavior. The aim of the integration of SFM into robot’s behavior is to increase robot’s ability to maintain its safety by avoiding the obstacles and collision with the other robots. The attractive feature of the proposed algorithm is the fact that the behavior of the humanoids will imitate the human’s behavior while avoiding the obstacle. The proposed algorithm combines formation control using Consensus Algorithm (CA) with collision and obstacle avoidance technique using SFM. Simulation and experiment results show the effectiveness of the proposed algorithm.

Developing a nonlinear adaptive control system for a fixed-wing unmanned aerial vehicle (UAV) requires a mathematical representation of the system dynamics analytically as a set of differential equations in the form of a strict-feedback systems. This paper presents a method for modeling a nonlinear flight dynamics of the fixed-wing UAV of BPPT Wulung in any conditions of the flight altitude and airspeed for the first step into designing a nonlinear adaptive controller. The model was formed into 10-DOF differential equations in the form of strict-feedback systems which separates the terms of elevator, aileron, rudder and throttle from the model. The model simulation results show the behavior of the flight dynamics of the Wulung UAV and also prove the compliance with the actual flight test results.

Rapid identification of content structures in a scientific paper is of great importance particularly for those who actively engage in frontier research. This paper presents a multi-classifier approach to identify such structures in terms of classification of rhetorical sentences in scientific papers. The idea behind this approach is based on an observation that no single classifier is the best performer for classifying all rhetorical categories of sentences. Therefore, our approach learns which classifiers are good at what categories, assign the classifiers for those categories and apply only the right classifier for classifying a given category. This paper employsk-fold cross validation over training data to obtain the category-classifier mapping and then re-learn the classification model of the corresponding classifier using full training data on that particular category. This approach has been evaluated for identifying sixteen different rhetorical categories on sentences collected from ACL-ARC paper collection. The experimental results show that the multi-classifier approach can significantly improve the classification performance over multi-label classifiers.

To improve the performance of phoneme based Automatic Speech Recognition (ASR) in noisy environment; we developed a new technique that could add robustness to clean phonemes features. These robust features are obtained from Complex Wavelet Packet Transform (CWPT) coefficients. Since the CWPT coefficients represent all different frequency bands of the input signal, decomposing the input signal into complete CWPT tree would also cover all frequencies involved in recognition process. For time overlapping signals with different frequency contents, e. g. phoneme signal with noises, its CWPT coefficients are the combination of CWPT coefficients of phoneme signal and CWPT coefficients of noises. The CWPT coefficients of phonemes signal would be changed according to frequency components contained in noises. Since the numbers of phonemes in every language are relatively small (limited) and already well known, one could easily derive principal component vectors from clean training dataset using Principal Component Analysis (PCA). These principal component vectors could be used then to add robustness and minimize noises effects in testing phase. Simulation results, using Alpha Numeric 4 (AN4) from Carnegie Mellon University and NOISEX-92 examples from Rice University, showed that this new technique could be used as features extractor that improves the robustness of phoneme based ASR systems in various adverse noisy conditions and still preserves the performance in clean environments.

In order to assist researchers in addressing time constraint and low relevance in using scientific articles, an automatic tailored multi-paper summarization (TMPS) is proposed. In this paper, we extend Teufel’s tailored summary to deal with multi-papers and more flexible representation of user information needs. Our TMPS extracts Rhetorical Document Profile (RDP) from each paper and presents a summary based on user information needs. Building Plan Language (BPLAN) is introduced as a formalization of Teufel’s building plan and used to represent summary specification, which is more flexible representation user information needs. Surface repair is embedded within the BPLAN for improving the readability of extractive summary. Our experiment shows that the average performance of RDP extraction module is 94.46%, which promises high quality of extracts for summary composition. Generality evaluation shows that our BPLAN is flexible enough in composing various forms of summary. Subjective evaluation provides evidence that surface repair operators can improve the resulting summary readability.

This paper investigates the problem of H∞ performance analysis for continous–time systems with two additive time-varying delays in the state. Our objective is focused on stability analysis of a continuous system with two time-varying delays with an H∞ disturbance attenuation level γ. By exploiting Lyapunov-Krasovski functional and introducing free weighting matrix variables, LMI stability condition have been derived.

Pemanfaatan twitter sebagai layanan customer serevice perusahaan sudah mulai banyak digunakan, tak terkecuali Speedy. Mekanisme yang ada saat ini untuk proses klasifikasi bentuk dan jenis keluhan serta informasi tentang jumlah keluhan lewat twitter masih dilakukan secara manual. Belum lagi data twitter yang bersifat tidak terstruktur tentunya akan menyulitkan untuk dilakukan analisa dan penggalian informasi dari data tersebut. Berdasarkan permasalahan tersebut, penelitian ini bertujuan untuk memproses data teks dari tweet pengguna twitteryang masuk ke akun @TelkomSpeedy untuk diolah menjadi informasi. Informasi tersebut nantinya digunakan untuk klasifikasi bentuk dan jenis keluhan. Merujuk pada beberapa penelitian terkait, salah satu metode klasifikasi yang paling baik untuk digunakan adalah metode Support Vector Machine (SVM). Konsep dari SVM dapat dijelaskan secara sederhana sebagai usaha mencari hyperplane yang dapat memisahkan dataset sesuai dengan kelasnya. Kelas yang digunakan dalam penelitian kali ini berdasarkan topik keluhan pelanggan yaitu billing, pemasangan/instalasi, putus (disconnect), dan lambat. Faktor penting lainnya dalam hal klasifikasi adalah penentuan feature atau atribut kata yang akan digunakan. Metode feature selection yang digunakan pada penlitian ini adalah term frequency (TF), document frequency (DF), information gain, dan chi-square. Pada penelitian ini juga dilakukan metode penggabungan feature yang telah dihasilkan dari beberapa metode feature selection sebelumnya. Dari hasil penelitian menunjukan bahwa SVM mampu melakukan klasifikasi keluhan dengan baik, hal ini dibuktikan dengan akurasi 82,50% untuk klasifikasi bentuk keluhan dan 86,67% untuk klasifikasi jenis keluhan. Sedangkan untuk kombinasi penggunaan feature dapat meningkatkan akurasi menjadi 83,33% untuk bentuk keluhan dan 89,17% untuk jenis keluhan. Kata Kunci—customer service, klasifikasi topik keluhan, penggabungan feature, support vector machine

Intrusion detection systems (IDS) play an important role in information security. Two major problems in the development of IDSs are the computational aspect and the architectural aspect. The computational or algorithmic problems include lacking ability of novel-attack detection and computation overload caused by large data traffic. The architectural problems are related to the communication between components of detection, including difficulties to overcome distributed and coordinated attacks because of the need of large amounts of distributed information and synchronization between detection components. This paper proposes a multi-agent architecture for a distributed intrusion detection system (DIDS) based on ant-colony clustering (ACC), for recognizing new and coordinated attacks, handling large data traffic, synchronization, co-operation between components without the presence of centralized computation, and good detection performance in real-time with immediate alarm notification. Feature selection based on principal component analysis (PCA) is used for dimensional reduction of NSL-KDD. Initial features are transformed to new features in smaller dimensions, where probing attacks (Ra-Probe) have a characteristic sign in their average value that is different from that of normal activity. Selection is based on the characteristics of these factors, resulting in a two-dimensional subset of the 75% data reduction.

Since abstract of scientific paper is author biased, readers’ required information may not be included in the abstract. Tailored summary may help them to get a summary based on their information needs. This research is the first one that implements tailored summary system for scientific paper. Tailored summary applies information extraction that transforms a scientific paper into Rhetorical Document Profile, a structured representation of paper content based on rhetorical scheme of fifteen slots. This research adapted building plan that used rhetorical scheme of seven slots. We also implement tailored summary system. After generating initial summary, surface repair is conducted to improve summary readability. Each sentence in initial summary is combined with template phrase based on syntax-tree combination method. There are five groups of template phrases provided in surface repair. We construct evaluation standards by asking five human raters. The best method for sentence selection subsystem that uses Maximal Marginal Importance-Multi Sentence is employing TF.IDF weighting system with precision/recall of 0.61. The surface repair subsystem has acceptance of 0.91.