IAES International Journal of Robotics and Automation (IJRA)
Not a member yet
460 research outputs found
Sort by
Multi-robot coverage algorithm in complex terrain based on improved bio-inspired neural network
Get PDFBiological neural network (BNN) algorithms have become popular in coverage search in recent years. However, its edge activity values are weak, and it is simple to fall into a local optimum at a late stage of coverage. When applied to complex environments, the 3D BNN network structure has high computational and storage complexity. In order to solve the above problems, we propose an algorithm for multi-robot cooperative coverage of complex terrain based on an improved BNN. The algorithm models the complex terrain using a 2.5-dimensional (2.5D) elevation map. Combining the dual-layer BNN network with the 2.5D elevation map, we propose an elevation value priority mechanism. This mechanism lets the robot make elevation-based decisions and prioritizes higher terrain areas. The dual neural network's first layer plans the robot's path in normal mode. The second network layer helps the robot escape the local optimum. Finally, the algorithm's full coverage effect in complex terrains and the speed of covering high terrain are verified by simulations. The experiments show that our algorithm preferentially covers high points of the region and eventually covers 100% of complex terrain. Compared with other algorithms, our algorithm covers more efficiently and takes fewer steps than others. The speed of covering high terrain areas has increased by 34.51%
Hybrid deep learning and active contour for segmenting hazy images
Get PDFImage segmentation seeks to distinguish the foreground from the background for further analysis. A recent study presented a new active contour model (ACM) for image segmentation, termed Gaussian regularization selective segmentation (GRSS). This interactive ACM is effective for segmenting certain objects in images. However, a weakness of the GRSS model becomes apparent when utilized on hazy images, as it is not intended for such conditions and produces inadequate outcomes. This paper introduces a new ACM for segmenting hazy images that hybridizes a pretrained deep learning model, namely DehazeNet, with the GRSS model. Specifically, the haze-free images are estimated using DehazeNet, which fuses the information with the GRSS model. The new formulation, designated as GRSS with DehazeNet (GDN), is addressed via the calculus of variations and executed in MATLAB software. The segmentation accuracy was evaluated by calculating Error, Jaccard, and Dice metrics, while efficiency was determined by measuring processing time. Despite the increased processing time, numerical experiments demonstrated that the GDN model achieved higher accuracy, as indicated by the lower error and higher Jaccard and Dice than the GRSS model. The GDN model can potentially be formulated in the vector-valued image domain in the future
Integration of natural language processing methods and machine learning model for malicious webpage detection based on web contents
Get PDFMalicious actors continually exploit vulnerabilities in web systems to distribute malware, launch phishing attacks, steal sensitive information, and perpetrate various forms of cybercrime. Traditional signature-based methods for detecting malicious webpages often struggle to keep pace with the rapid evolution of malware and cyber threats. As a result, there is a growing demand for more advanced and proactive approaches that can effectively identify malicious web content based on its characteristics and behavior. Detection based on web content is crucial because malicious webpages can be designed to mimic legitimate ones, making them difficult to identify through traditional means. By analyzing the content of webpages, it becomes possible to uncover patterns, anomalies, and malicious intent that may not be evident from surface-level inspection. The proposed approach integrates a pretrained Word2Vec model with seven distinct machine learning classifiers to enhance malicious webpage detection. Initially, web contents (documents) are encoded using the Word2Vec model, followed by the computation of average Word2Vec embeddings for each document. Subsequently, each classifier is trained on the extracted average Word2Vec embedding features. The results demonstrate that the Word2Vec model significantly enhances the detection accuracy, achieving an accuracy of 94.8% and an F1-score of 94.9% with the random forest classifier, and an accuracy of 94.6% and an F1-score of 94.7% with the extreme gradient boosting classifier
Advanced cardiovascular disease classification using multi-modal imaging and deep learning
Get PDFCardiovascular disease (CVD) is a disorder of the heart and blood vessels that causes significant morbidity and mortality. They also represent a global public health burden and the primary cause of death worldwide. In this research, a novel deep learning-based multi-model image (DL-MMI) has been proposed for detecting CVD. Initially, the input Kaggle datasets images like magnetic resonance imaging (MRI), computed tomography (CT), positron emission tomography (PET), and chest X-ray are fed into wavelet transform-based Multiscale Retinex in the pre-processing phase to enhance the quality of the images. Then the enhanced images are given to GLCM for extracting features in the images. Finally, the dilated convolutional neural network (D-CNN) is used to classify healthy and CVD images. The experimental findings use the specific measures of accuracy, recall, precision, specificity, and F1-score to demonstrate the durability of the DL-MMI approach. Using the Kaggle dataset the proposed DL-MMI method achieves an accuracy rate of 98.89%. The proposed DL-MMI model increases the overall accuracy by 28.62%, 7.51%, and 17.57% than the existing methods such as convolutional auto encoder, CNN, and deep learning, respectively
Humanoid robot balance control system during backward walking using linear quadratic regulator
Get PDFHumanoid robots are designed to replicate human activities, including tasks in hazardous environments. However, maintaining balance during backward walking remains a significant challenge due to center of mass (CoM) shifts beyond the support polygon and limited knee joint motion. This study proposes a control strategy that integrates a linear quadratic regulator (LQR) with optimized walking patterns to enhance dynamic stability. The approach combines LQR-based control with CoM trajectory planning to ensure safe and stable backward walking. The methodology includes inverse kinematics for generating walking patterns and the use of Inertial Measurement Unit (IMU) sensors to estimate the CoM trajectory. LQR parameters were tuned through simulation to improve responsiveness to disturbances. Evaluation metrics focused on CoM deviation, rise time, settling time, and overshoot. Experimental results demonstrate that the proposed LQR system effectively maintains the CoM within 5% of the support polygon boundary. The system achieved rise times under one second and settling times below two seconds, while minimizing pitch and roll overshoots. Compared to proportional control, the proposed method significantly improves stability and reduces the risk of falling. This research advances control strategies for humanoid robots, contributing to improved mobility and operational safety. Moreover, it supports Sustainable Development Goal (SDG) 9 by promoting innovation in intelligent robotic systems that can assist in complex or high-risk environments
Design of H-/H∞ based fault detection filter for linear uncertain systems using linear matrix inequalities
Get PDFOne of the significant challenges in model-based fault detection is achieving robustness against disturbances and model uncertainties while ensuring sensitivity to faults. This study proposes an optimized approach for designing fault detection filters for discrete-time linear systems with norm-bounded model uncertainties. The design leverages the H-/H∞ optimization framework and is expressed through linear matrix inequality constraints. The filter is designed to produce a residual signal that balances two opposing objectives: minimizing the impact of disturbances and model uncertainties while maximizing fault sensitivity. The effectiveness of the proposed method is demonstrated through simulations involving sensor and actuator fault detection in the well-known three-tank system. Simulation results illustrate the method's ability to maintain robustness against disturbances and uncertainties while effectively detecting faults in the system
A method integral sliding mode control to minimize chattering in sliding mode control of robot manipulator
Get PDFThis paper presents an improved sliding mode control (SMC) strategy for robotic manipulators by introducing a novel exponential integral-based adaptive gain law, referred to as integral sliding mode control (ISMC). The proposed approach dynamically adjusts the switching gain KKK in real-time, based on the accumulated system error, thereby effectively reducing chattering while preserving system robustness. Unlike many existing methods, the ISMC strategy eliminates the need for state observers or complex estimation techniques, simplifying implementation. Theoretical analysis is provided using Lyapunov stability theory, ensuring global convergence. Simulation results on 2-DOF and 3-DOF robotic arms demonstrate superior tracking accuracy and smoother control signals compared to conventional SMC approaches. This work contributes a lightweight yet effective SMC enhancement with practical benefits for real-world robotic applications
Design and implementation of QUADRESCUE: A ROS-based quadruped robot for disaster response support
Get PDFSearch and rescue (SAR) operations in hazardous environments demand robotic systems capable of traversing complex terrains while ensuring responder safety. Traditional wheeled platforms often fail in debris-laden areas, and fully autonomous quadrupeds remain financially out of reach for many rescue agencies. This paper presents the design and development of QUADRESCUE, a modular operator-assisted quadruped robot built to bridge the gap between affordability and capability in disaster response. QUADRESCUE delivers core SAR functionalities including remote visual inspection, real-time terrain mapping via an RGB-D camera, payload transport, and GPS-based survivor localization. Built with a robust three degrees of freedom (3DoF) per leg design, the robot uses inverse kinematics algorithms to precisely control twelve servo motors for stable locomotion across uneven terrain. The system integrates the robot operating system (ROS) for seamless operation, real-time joystick control for easy navigation, an IMU for orientation sensing, and a GPS module with 3-meter accuracy. Field evaluations demonstrate 80–94% success rates on challenging surfaces, substantially outperforming wheeled counterparts 19% to 39% with a 200-meter control range and 45 minutes of runtime. QUADRESCUE offers a lightweight, cost-effective, and repairable solution that combines practical usability with advanced performance, making it well-suited for real-world deployment in emergency rescue situations
Optimizing robot anomaly detection through stochastic differential approximation and Brownian motion
Get PDFThis paper presents an adaptive approximation method for detecting anomalous patterns in extensive data streams gathered by mobile robots operating in rough terrain. Detecting anomalies in such dynamic environments poses a significant challenge, as it requires continuous monitoring and adjustment of robot movement, which can be resource intensive. To address this, a cost-effective solution is proposed that incorporates a threshold mechanism to track transitions between different regions of the data stream. The approach utilizes stochastic differential approximation (SDA) and optimistic optimization of Brownian motion to determine optimal parameter values and thresholds, ensuring efficient anomaly detection. This method focuses on minimizing the movement cost of the robots while maintaining accuracy in anomaly identification. By applying this technique, robots can dynamically adjust their movements in response to changes in the data stream, reducing operational expenses. Moreover, the temporal performance of the data stream is prioritized, a key factor often overlooked by conventional search engines. This paper demonstrates how the approach enhances the precision of anomaly detection in resource-constrained environments, making it particularly beneficial for real-time applications in rugged terrains
Localization and mapping of autonomous wheel mobile robot using Google cartographer
Get PDFCOVID-19 has become a world concern because of the spread and number of cases that have befallen the world. Medical workers are the first exposed group because they have direct contact with patients. So, a vehicle is needed to replace tasks such as logistics, delivery, and patient waste transportation. An autonomous wheeled mobile robot (AWMR) is a wheeled robot capable of moving freely from one place to another. AWMR is required to have good navigation and trajectory control skills. The purpose of this study is to develop an AWMR navigation system model based on the simultaneous localization and mapping (SLAM) algorithm, accurately in a dynamic environment. With this research, developing a good navigation and trajectory method for AWMR, in the future, it can be applied to produce an AWMR platform for multipurpose. This research was conducted in two stages of development. The first year is the research that is currently being carried out, focused on sensor modeling, designing SLAM-based navigation models, and making navigation system testbeds. This research produces a trajectory navigation and control system that can be implemented on an AWMR platform for the purposes of logistics, transportation, and patient waste in hospitals