Journal of Robotics and Control (JRC)

Enhanced RRT* with APF and Halton Sequence for Robot Path Planning

Mohammed T. Hameed — 2025-03-04

This paper presents a new path planning method (APF-IRRT*-HS), which relies on the optimization process of the conventional RRT* algorithm and combined with the APF method where the sampling process of the RRT* algorithm is improved using the Halton sequence, which is known to be deterministic and repeatable and provides more efficient coverage than other low discrepancy sequences with the modified goal-based method which provides a probabilistic approach to decide whether to sample from a point directly at the target or to choose a random point from the Halton sequence based on the current distance. We implemented the proposed method in two cases of mass point and two-link robots. The proposed method compares path length with the conventional RRT* algorithm and APF-RRT*, as well as time efficiency and number of iterations. The technique proves effective in various dynamic environments. Specifically, the APF-IRRT*-HS algorithm achieved an improvement of approximately 21.88% and 7.5% in path length, 79.75% and 49.2% in computation time, and 57.39% and 40% in the number of iterations compared with the RRT* and RRT*-APF algorithms, respectively. We can use this method in everyday applications such as robotic arms, drones, self-driving cars, etc. More advanced methods, such as multi-link robots and real-time constraints, can be used in the future.

Image Denoising Using Generative Adversarial Network by Recursive Residual Group

Maysaa A. Ulkareem Naser — 2025-03-05

Cardiac magnetic resonance imaging (CMR) is a vital tool for noninvasively assessing heart shape and function, offering exceptional spatial and temporal resolution alongside superior soft tissue contrast. However, CMR images often suffer from noise and artifacts due to cardiac and respiratory motion or patient movement impacting diagnostic accuracy. While real-time noise suppression can mitigate these issues, it comes at a high computational and financial cost. This paper introduces a method that includes a complete way to clean up medical images by using a new Denoising Generative Adversarial Network (D-GAN). The D-GAN architecture incorporates a recursive residual group-based generator and a discriminator inspired by PatchGAN.The recursive residual group-based generator and the Selective Kernel Feature Fusion (SKFF) mechanism are part of a new D-GAN architecture that makes denoising work better. A PatchGAN-based discriminator designed to improve adversarial training dynamics and texture modeling for medical images. These innovations offer improved feature refinement and texture modeling, enhancing the denoising of cardiac MRI images. allows the model to get a doubling context of local and global, informational, and hierarchical developed features located in the generator. Our technique outperforms other methods in terms of PSNR and SSIM. With scores of 0.837, 0.911, and 0.971 for noise levels of 0.3, 0.2, and 0.1, and PSNR scores of 29.48 dB, 32.58 dB, and 37.85 dB, the results show that the D-GAN method is better than other methods.

Design of a Robust Component-wise Sliding Mode Controller for a Two-Link Manipulator

Mohammed Qasim — 2025-03-08

Compared to conventional Multiple-Input Multiple-Output (MIMO) Sliding Mode Control (SMC) techniques, the component-wise SMC approach offers several advantages, including improved decoupling of system dynamics, enhanced robustness, and greater flexibility in controller design. This paper proposes a novel trajectory tracking controller for a two-link manipulator based on the component-wise sliding mode control approach. The design methodology involves determining controller gains by solving a set of inequalities. This analysis results in conditions on the system parameter uncertainties that guarantee the existence of a feasible solution to the set of inequalities. Furthermore, an algorithm is presented to determine the maximum allowable uncertainties that ensure the feasibility of the controller gains. To evaluate the performance and robustness of the proposed tracking controller, the manipulator is subjected to a series of challenging trajectories, including circular and figure-8 ones, under both nominal and maximum allowable uncertainty conditions. The proposed controller demonstrates superior performance across both circular and figure-8 trajectories, exhibiting excellent transient response and minimal steady-state error even under the maximum permissible uncertainties, which extend up to 27% in link masses. This performance is validated through a quantitative analysis that incorporates a comparative evaluation against two conventional MIMO SMC techniques. The comparison is conducted using the Integral Norm of Error (INE) to assess tracking accuracy and the Integral Norm of Control Action (INU) to evaluate the energy efficiency of the controllers. These metrics provide a comprehensive basis for analyzing both the precision and the energy consumption of the proposed control strategy in relation to established methods.

Hybrid SVD and SURF-Based Framework for Robust Image Forgery Detection and Object Localization

Fallah H. Najjar — 2025-03-10

This paper presents a highly effective and reliable approach for detecting image forgery and identifying manipulated regions in digital images. The proposed method uses a combination of Singular Value Decomposition (SVD) and the Speeded-Up Robust Features (SURF) algorithm, achieving a high degree accuracy of 99.1% for revealed tampering. After an input image is initially divided parallel to partition, then is performed by SVD to extract features with remarkable discriminability, the method is valued based on independent experiments. The norms are calculated, and pixels with the same norm begin to group to identify potentially tampered areas. In order to simplify the detection process, we conduct a weighted comparison among subgroups to distinguish real structures from false ones. Once we discover a suspicious forgery area, the SURF algorithm comes into play to accurately identify the manipulated items. This process uses a keypoint detector, descriptor calculations, the match between points, and geometric checking to improve the accuracy and reliability of forgery localization. Experimental results on different image databases show that this method is effective. It exhibits advanced ability in detecting forgeries, finding objects and locating where they are in an image. Eventually, we hope this work will produce a sturdy forgery detection system and improve the accuracy of recognizing tampered regions. The proposed method is useful in digital forensics and image verification.

Optimizing Mobile Robot Path Planning with a Hybrid Crocodile Hunting and Falcon Optimization Algorithm

Wassan Adnan Hashim — 2025-03-10

Thorough path planning is critical in unmanned ground vehicle control to reduce path length, computational time, and the number of collisions. This paper aims to introduce a new metaheuristic method called the Hybrid Crocodile Hunting-SearcH and Falcon Optimization (CHS-FO) algorithm. This method combines CHS's exploration and exploitation abilities with FO's rapid convergence rate. In this way, the use of both metaheuristic techniques limits the disadvantage of the individual approach, guaranteeing a high level of both global and local search. We conduct several simulations to compare the performance of the CHS-FO algorithm with conventional algorithms such as A* and Genetic Algorithms (GA). It is found The results show that the CHS-FO algorithm performs 30–50% better in terms of computation time, involves shorter path planning, and improves obstacle avoidance. Eristic also suggests that the path generation algorithm can adapt to environmental constraints and be used in real-world scenarios, such as automating product movement in a warehouse or conducting search and rescue operations for lost vehicles. The primary The proposed CHS-FO architecture makes the robot more independent and better at making choices, which makes it a good choice for developing the next generation of mobile robotic platforms. Goals will encompass the improvement of the algorithm's scalability for use in multiple robots, as well as the integration of the algorithm in a real environment in real time.

Optimizing PID Controller for Large-Scale MIMO Systems Using Flower Pollination Algorithm

Suad Ali Aessa — 2025-03-11

In communications systems, a group of technologies can be linked into one system. Each technology has a function, and each system has stages. Therefore, it can be said that the input stage can have one or multiple inputs. MIMO techniques suffer from a large-scale linear dynamic problem, it will be easy to adjust the (PID) of a continuous system, and any system is considered vulnerable to disturbances during the operation process. Therefore, a state of instability can occur in it, which requires developing solutions to modify the behavior of the it. Systems need control units to handle transient states as a result of changing operating conditions of the system. Expert and intelligent systems can be used to adjust the traditional controllers and make them adapt to the operating conditions of the proposed system. Work must be done to make the proposed approach capable of ensuring stability for the system. Work can be done to reduce the time for the transient state and the speed of response to the stable state. The behavior of the system can be clarified through simulation results and show the difference between the methods. Proposed to test the feasibility and effectiveness of the work and verify it using the MATLAB program to design a highly accurate and efficient model. The current study reviews This work displays the tuning of the PID controller for MIMO systems utilizing a statistical FPA and evaluated by objective function as integral time absolute error (ITAE). A combination of ITAE combined with the FPA reduction method is adopted to reduce the steady-state transient time responses between the higher-order initial scheme and the unit amplitude It also aims to conduct simulation and develop the appropriate and proposed design model with different. It is possible to compare the control of the system using a traditional control unit and another that adjusts using the modern technology of the Flower pollination algorithm FPA-PID. It also showed that the results of the simulation process were clear that the optimization process using FPA-PID was superior to the other traditional case (PID).

Improving Short-Term Electrical Load Forecasting with Dilated Convolutional Neural Networks: A Comparative Analysis

Tuan Anh Nguyen — 2025-03-12

Short-term load forecasting (STLF) is vital for grid stability and resource optimization for energy systems. Accurate forecasting helps maintain a stable power supply, reduce costs, and improve decision-making. Traditional convolutional neural networks (CNNs) capture local patterns well but struggle with long-term dependencies under fluctuating conditions. This study introduces an optimized Dilated Convolutional Neural Network (DCNN) to enhance accuracy in short- and long-term load forecasting. The key contribution is a new DCNN framework that expands the receptive field without adding computational complexity, effectively capturing multi-level temporal dependencies. This improves performance, stability, and accuracy in volatile conditions. The methodology applies dilated convolution techniques to a real-world electricity load dataset with 13,440 hourly data points. Preprocessing includes normalization and outlier removal. Hyperparameter tuning optimizes dilation rates, kernel sizes, and learning rates. Results show that the DCNN outperforms traditional models, achieving the lowest Mean Absolute Percentage Error (MAPE) of 0.0096. These results surpass CNN (MAPE: 0.0116), GRU (MAPE: 0.0102), and Long Short-Term Memory (LSTM) (MAPE: 0.0272) models. The DCNN also maintains efficiency and stability with volatile data. In conclusion, optimized dilated convolution techniques significantly enhance load forecasting, offering scalable, robust solutions for modern energy management systems requiring fast, accurate, and reliable predictions.

Adaptive Intrusion Detection for IoT Networks using Artificial Immune System Techniques: A Comparative Study

Amaal Rateb Shorman — 2025-03-13

The rapid proliferation of IoT devices has led to a significant increase in security vulnerabilities, rendering them susceptible to more sophisticated assaults. Conventional security methods often encounter difficulties in the changing surroundings and resource limitations of IoT, requiring flexible, low-resource alternatives. This research proposes the use of three distinct Artificial Immune System (AIS) methodologies to enhance the security of the Internet of Things (IoT). The concepts include clonal selection, negative selection, and risk theory. Each algorithm fulfills essential security requirements: Negative selection helps find new dangers, clonal selection finds things that aren't normal in real-time, and risk theory uses context-aware responses to reduce false positives. When tested on several IoT-specific datasets, the AIS framework had an average detection accuracy of 94%. It also had a 20% reduction in false-positive rates and made better use of resources than traditional machine learning models like SVM, RF, and KNN. The findings indicate that the framework is effective for resource-constrained IoT devices. They enhance IoT security by using adaptive, immune-inspired countermeasures tailored to the unique problems of IoT. The suggested approach guarantees that networked devices remain adequately protected against new threats. The conclusions indicated that integrating comprehensive security management into IoT frameworks might markedly diminish total risk, therefore facilitating safer and more dependable IoT applications.

Proton Exchange Membrane Fuel Cell Combined with Battery and Flywheel Energy Storage for Sustainable Power and Clean Electric Trike Vehicle

Ganesha Tri Chandrasa — 2025-03-14

An electric personal three-wheeler, referred to as the Reverse Trike Electric Vehicle (RTEV), has been developed as a forward-looking solution for clean transportation within communities. This innovative vehicle integrates multiple energy storage systems, including the mechanical energy of a flywheel, the chemical energy of a hydrogen (H2) fuel cell system, and the chemical energy stored in a lithium-ion battery. The vehicle is powered by a 500-watt, 48-volt platform that incorporates a DC brushless electric motor, a battery, and a fuel-cell. Hydrogen is stored in two steel tube containers filled with ultra-high purity (UHP) hydrogen, with output pressure ranging from 10 to 35 psi. The lithium battery pack features a configuration of 13 series and 3 parallel (13S3P) cells, assembled using 18650-sized lithium-ion cells. Additionally, two flywheels of varying mass were utilized in the design. The prototype underwent testing both indoors on a dynamometer test bed and outdoors on community roads for analysis. The results from these tests clearly demonstrate the contributions of each energy storage system to the vehicle's power traction and distance performance, showcasing the effectiveness of this multi-faceted approach to clean transportation.

A Comparative Study of Nonlinear Control and Passivity-Based Control using Neural Networks for A Bicycle Robot

Minh Ngoc Huynh — 2025-03-14

In this paper, a comparative study of nonlinear control and passivity-based control using neural networks for a bicycle robot is proposed. Bicycle robot is a nonlinear, multi-input multi-output system. Two inputs of a bicycle robot are the steering torque and kinetic energy. Its two outputs are the steering angle and the rolling angle. The control problem is that the steering angle and the rolling angle track a value of zero, and the velocity of the steering angle and velocity of the rolling angle track a value of zero to make a bicycle robot stabilize at its vertical balance. Firstly, an input-output linearization control law decouples the bicycle robot into single-input single-output systems. This plant is passive and zero-state observable. Secondly, the passivity-based control law is applied to each single-input single-output system. Finally, the neural network, which performs the passivity-based control, is applied to each single-input single-output system in order that the bicycle robot keeps its vertical balance. A training algorithm using the steepest descend method is proposed. The simulation results of the passivity-based control and the results of the passivity-based control using neural networks show that the bicycle robot keeps its vertical balance. The settling time of the steering angle and the rolling angle of the passivity-based control using a neural network, 1.8s, is shorter than that of the passivity-based control. There is a comparison with the passivity-based control combined with sliding mode control for a bicycle robot.

Analysis and Design of a Robust Security System to Address Cybersecurity Vulnerabilities and Control the Security of the Internet of Vehicles

Rafah Kareem Mahmood — 2025-03-17

The Internet of Vehicles (IoV) system faces some security vulnerabilities, which could lead to cyberattacks targeting smart vehicles. This paper investigates the cybersecurity challenges of IoV, focusing on defending smart cars from covert attackers. The primary aim of the study is to design a robust security system to combat and prevent these cyber-attacks. The work makes two main contributions: First, it explores the security vulnerabilities in inter-vehicle communication, particularly focusing on two types of cyber attacks. To address these vulnerabilities, an intelligent detection algorithm is proposed to close security gaps within the system. The study also discusses the role of smart vehicles and their interconnected systems, which communicate through the Controller Area Network (CAN) bus and other external communication units. Second, the study emphasizes the importance of protecting critical vehicle systems such as the engine control unit (ECU), adaptive cruise control (ACC), anti-lock braking system (ABS), and central locking system. These systems are crucial for vehicle safety, and any compromise could have disastrous consequences. The proposed approach utilizes a Proportional-Integral-Derivative (PID) controller to monitor and control acceleration, leveraging real-time measurements of speed and distance to prevent accidents and mitigate security risks. The results demonstrate the ffectiveness of the proposed solution in securing smart vehicles from cyber threats, by ensuring safe dynamic vehicle systems under potential cyber-attacks.

Development and Evaluation of a Low-Cost CNC Wood Carving Machine for Artisanal Applications

Anees Abu Sneineh — 2025-03-18

Improving the quality of the produced artwork would require the development of computer numerical control (CNC) technology for various wood shaping processes. This paper presents the development and implementation of a CNC wood carving machine. This project aims to produce wood carvings with high precision and low cost. The electronic components were selected, the mechanical frame of the machine was built, and all parts of the machine were assembled and installed. The presented wood carving machine is based on the principle of drawing the desired shape, translating it into G-code, and then sending it to the microcontroller program. The microcontroller sends instructions to the CNC shield, which drives three stepper motors in a synchronized manner in order to produce the desired carving model. The maximum workpiece of the caving machine designed to be 30x30 cm. Experiments were conducted to test the functionality of the proposed CNC wood carving machine. The results showed that the carved models created by this machine had a carving precision range from ±0.1 to ±0.15 mm and a carving speed of approximately 500 rpm. The proposed CNC can be used to produce low-cost artisanal woodwork applications.

Revolutionizing Numerical Approximations: A Novel Higher-Order Implicit Method vs. Runge-Kutta for Initial Value Problems

Mohammad W. Alomari — 2025-03-18

This work is dedicated to advancing the approximation of initial value problems through the introduction of an innovative and superior method inspired by Taylor’s approach. Specifically, we present an enhanced variant achieved by accelerating the expansion of the Obreschkoff formula. This results in a higher-order implicit corrected method that outperforms Rung– Kutta’s (RK) method in terms of accuracy. We derive an error bound for the Obreschkoff higher-order method, showcasing its stability, convergence, and greater efficiency than the conventional RK method. To substantiate our claims, numerical experiments are provided, highlighting the exceptional efficacy of our proposed method over the traditional RK method.

Adaptive Neural Network-Based Voltage Regulation for a High-Gain Boost Converter in Solar Photovoltaic Systems

Mohcine Byar — 2025-03-18

This study proposes an adaptive Artificial Neural Network-based voltage control strategy for maintaining a stable DC bus voltage in a high-gain DC-DC boost converter for solar photovoltaic systems. Unlike conventional PID controllers, which struggle with non-linear and dynamic conditions, the proposed controller dynamically adjusts the duty cycle to mitigate the effects of varying solar irradiance and reference voltage, ensuring robust voltage regulation with reduced overshoot, enhanced transient response, and improved steady-state stability. This approach addresses critical challenges in standalone solar applications, such as water pumping and rural electrification, where consistent performance is essential despite fluctuating environmental conditions. In comparison to conventional control strategies, the ANN-based controller demonstrates superior adaptability, particularly under rapidly changing operating conditions. The results demonstrate the superior adaptability and efficiency of the ANNbased controller compared to the conventional PID controller, making it a valuable and reliable solution for sustainable solar PV systems. The proposed system was validated using a cosimulation framework that integrates MATLAB/Simulink and OrCAD, facilitating performance evaluation under varying solar irradiance and reference voltage conditions.

Secure and Efficient Mutual Authentication Protocol for VANETs Using Edge Computing and Signature-Based Cryptography

Abdulnasser AbdulJabbar Abbood — 2025-03-21

Security in vehicular ad-hoc networks (VANETs) as a result, vehicular ad-hoc networks (VANETs) need to adopt and implement strong security protocols respective to vehicle-tovehicle and vehicle-to-infrastructure communication. But state-ofthe-art authentication methods have drawbacks like computational overhead, scalability issues, and susceptibility to identity stealing, replay attacks and data manipulation. To mitigate these issues, we present an innovative protocol for mutual authentication in edge computing assisted VANEts by employing an elliptic curve signature-based to improve the security and performance of the protocol. The proposed scheme guarantees low-latency authenticated by offloading computation tasks to edge nodes and simultaneously provides conditional privacy-preserving vehicle tracing for law enforcement. Formal security verification using ProVerif shows to be resilient towards replays, man-in-the-middle and eavesdropping attacks. Simulation results also show that the proposed protocol achieves highly efficient computational and communication overhead in comparison with current approaches. The performance results are promising and therefore, the proposed scheme can I be considered as practical and scalable for realistic applications in VANET.

A Novel Integral Interconnection Damping Assignment Passivity-Based Control Approach for Underactuated Inverted Pendulum System

Minh-Duc Tran — 2025-03-21

This study explores an application of robust nonlinear control to an underactuated inverted pendulum system (UIPS), a type of underactuated mechanical system, by first transforming the perturbed system into a form of a port-controlled Hamiltonian system (PCH), then utilizing the Interconnection and Damping Assignment Passivity-Based Control (IDA-PBC) methodology to achieve the stabilization of an unperturbed closed-loop PCH system with an assigned energy function that qualifies as a Lyapunov candidate at the unstable equilibrium point. From there, the problem of robustification of IDA-PBC for a perturbed closed-loop PCH system with the state-dependent input matrix of the UIPS, subject to constant matched disturbances, is addressed by adding an outer-loop controller with an additional state. This results in a new system that preserves the framework of a PCH system, rejects the disturbance, and has a new energy function that again serves as a Lyapunov function at the desired equilibrium point. This proposed methodology is called integral IDA-PBC (iIDA-PBC). The effectiveness and applicability of the proposed method are thoroughly assessed through numerical simulations and experimental validation on the UIPS. The results demonstrate the method's proficiency in handling the system’s constant matched disturbances and model inaccuracies, underscoring the potential of iIDA-PBC for broader applications in systems facing similar control challenges.

Synthesis of Active Disturbance Rejection Controller via Extended State Observer Combined with LQR Controller for Two-Wheeled Line Tracking Robot

Nguyen Xuan Chiem — 2025-03-21

This paper presents a method of synthesizing control laws based on the LQR controller and ADRC method for a two-wheel differential line-following robot when the robot dynamics have uncertain factors. First, the mathematical model includes line-following kinematic and dynamic models. LQR controller is designed based on the linear model of the robot when coincident with the line. When the robot has uncertain factors such as model parameter uncertainty and impact noise, the LQR controller will not ensure the control quality of the system. To overcome this, two observers are designed to observe the linear velocity and angular velocity states of the robot. This ensures more complete and accurate information of the model states in the LQR control law. The effectiveness of the control law is demonstrated through numerical simulation results and compared with the LQR controller.

Comparison and Evaluation of Stability-Preserving Model Order Reduction Methods for Rigid Robot Manipulators: A Study on a 4-DOF Robotic Arm

Yen-Vu Thi — 2025-03-22

Robust control of robotic systems critically depends on the stability-preserving capabilities of model order reduction (MOR) techniques. However, selecting an optimal MOR method for rigid robot manipulators remains challenging due to trade-offs between model fidelity, stability preservation, and computational efficiency. The research contribution of this study is to systematically compare three MOR methods—Balanced Truncation (BT), Positive-Real Balanced Truncation (PRBT), and Modal Truncation (MT)—applied to a 4-degree-of-freedom (4-DOF) robotic arm modeled as a linear time-invariant (LTI) system. We evaluated the methods based on error metrics, including H-infinity norm differences, and analyzed their time-domain and frequency-domain responses under standard test conditions. Our results demonstrate that BT provides superior reduction quality by maintaining stability and achieving an accurate dynamic response. PRBT, while exhibiting higher error than BT, effectively preserves both stability and passivity, making it advantageous for applications where passivity is essential, such as in mechanical and electrical circuits. In contrast, MT shows significant performance limitations with large errors and inconsistent responses, rendering it unsuitable for precision control applications. In conclusion, this study offers valuable insights into the trade-offs among MOR techniques and highlights practical implications for industrial automation. Future work will focus on expanding the analysis to a broader range of robotic systems and varying operational conditions.

Advanced Cybersecurity Framework for LEO Aerospace: Integrating Quantum Cryptography, Artificial Intelligence Anomaly Detection, and Blockchain Technology

Makhabbat Bakyt — 2025-03-24

This study aims to enhance the security of high-speed Low Earth Orbit (LEO) communication systems by developing an integrated, multi-layered security framework that addresses the limitations of current aerospace cybersecurity measures. The primary challenge lies in ensuring real-time data confidentiality, integrity, and authenticity in the face of sophisticated quantum and spoofing threats. To overcome these issues, the research contribution is the design and evaluation of a unified framework that combines quantum-resistant encryption using a FACT system with a Quantis USB quantum random number generator, an LSTM encoder-decoder model for real-time anomaly detection in ADS-B messages, and a blockchain-based mechanism for immutable data logging. The methodology involves benchmarking quantum-enhanced AES against traditional encryption schemes, training the LSTM network to detect subtle anomalies in flight data, and assessing blockchain scalability under high transaction loads. Results indicate significant improvements in encryption speed and detection accuracy—demonstrating up to a 30% increase in anomaly detection performance—while also revealing challenges such as increased computational overhead and scalability limitations in blockchain implementation. The framework shows promise for practical applications in satellite communications and air traffic management, though further research is needed to optimize resource consumption and enhance system resilience under extreme operational conditions.

Improved Path Planning for Multi-Robot Systems Using a Hybrid Probabilistic Roadmap and Genetic Algorithm Approach

Thanushika Jathunga — 2025-03-24

This study focuses on the development and application of an improved Probabilistic Roadmap (PRM) algorithm enhanced with Genetic Algorithms (GA) for multi-robot path planning in dynamic environments. Traditional PRM-based methods often struggle with optimizing path length and minimizing turns, particularly in complex, multi-agent scenarios. To address these limitations, we propose a hybrid PRM-GA approach that incorporates genetic operators to evolve optimal paths for multiple robots in real-time.The research contribution is an enhanced PRM-GA framework that improves efficiency in multi-robot navigation by integrating evolutionary techniques for dynamic obstacle handling and optimized path generation.The research methodology involves testing the algorithm in various environments, including varying robot numbers and environmental complexities, to evaluate its scalability and effectiveness. Our results demonstrate that the PRM-GA algorithm successfully reduces both path lengths and turn counts compared to standard PRM-based methods, ensuring collision-free and smooth paths. The algorithm showed robust performance across different scenarios, effectively handling dynamic obstacles and multi-agent coordination. However, in highly dynamic environments with rapidly changing obstacles and constraints, the algorithm may occasionally produce paths with turn counts and distances similar to or slightly higher than those of simpler approaches due to the need for frequent re-optimization. Future research can explore incorporating additional factors such as energy consumption and time optimization, alongside distance and turns, to further enhance the algorithm's efficiency in real-world applications. Overall, the PRM-GA approach advances the state of the art by offering a more adaptable and scalable solution for multi-robot path planning, with applications in logistics, industrial automation, and autonomous robotics.

Applications of Robotic Systems and Emerging Technologies in the Collection, Classification, and Treatment of Medical Waste: A Systematic Review

Roger Fernando Asto Bonifacio — 2025-03-25

The growing volume of infectious medical waste represents a critical challenge for public health and environmental protection, demanding more efficient solutions in hospital waste management (MWM). The contribution of this research is to offer a systematic review of advances in robotic systems and emerging technologies, including artificial intelligence, blockchain, Internet of Things (IoT), and drones applied to the collection, classification, and treatment of medical waste. To do this, a search strategy was used in databases (Scopus, Web of Science, IEEE Xplore, among others) between 2019 and 2024, using keywords related to "medical waste," "robotics," "AI," and "blockchain." After applying inclusion and exclusion criteria, 107 studies were selected that report at least partial validations in health settings. The results show that integrating robotic systems with computer vision and deep learning algorithms can improve classification accuracy by more than 90% while reducing the direct exposure of personnel to infectious material. In addition, technologies such as blockchain and IoT strengthen data traceability and security, although economic and regulatory challenges that limit their large-scale adoption persist. In conclusion, robotics and emerging technologies constitute a promising approach to optimize medical waste management. Even so, greater standardization of performance metrics and validations in high-capacity hospitals are required, as well as the active involvement of health authorities to promote their implementation.

Three-Dimensional Object Detection in Point Clouds with Multi-Stage Proposal Refinement Network

Jyothsna Datti — 2025-03-28

Three-dimensional object detection in point clouds serves a vital role in autonomous driving and robotics. Point Clouds provide a vivid representation of 3D data that enables reliable object detection by acquiring the spatial distribution of points in a scene, facilitating the localization and identification of the objects within three-dimensional space. Precise localization of the objects remains challenging, particularly for moderately visible objects which attributes to inconsistent quality proposals. To tackle this, this paper presents a multi-stage proposal refinement network to generate the qualitative predictions. The research contribution is, first to improve the quality of proposals in partially visible objects, the model is integrated with 3D Resnet backbone through the refinement module at various stages. Second, to improve the quality of predictions, a confidence-weighted box voting mechanism is incorporated ensuring the precise bounding box detections. Experimentation analysis was carried out on the KITTI, NuScenes and the custom LIDAR datasets. Notably, the proposed method achieves an average precision of 82.45% for Car class, 44.94% for Pedestrian class and 66.12% for Cyclist class on the moderate category of KITTI dataset, but in the hard category with high occlusion need to be improved. On Nuscenes dataset, the model achieved mAP of 66.2%. In custom dataset, 2739 training frames, 342 frames for validation, and 343 frames for testing were taken which achieved an average precision of 82.40% for Car, 44.10% for pedestrian and 67.90% for Cyclist. The results indicate that multi-stage refinement network enhances to perform the object detection precisely, which is critical to localize and detect the target in autonomous driving and robotics.

Bio-Inspired Robotics: Kinematic and Gait Analysis of Quad and Hexa-Legged Systems

Spoorthi Singh — 2025-04-08

Navigating hazardous environments, such as areas with fire risks, wild animal activity, or inaccessible terrains, poses significant challenges, necessitating the development of bio-inspired robotic systems. This study focuses on the biomechanical design and kinematic analysis of a spider-mimicking robot, specifically examining quad and hexa-legged configurations to optimize movement efficiency and stability. The research employed 3D Computer-Aided Design (CAD) in Fusion 360 to model and simulate the robot's leg framework, analyzing deformation, tension, and strain. Fused Deposition Modelling (FDM) with Poly Lactic Acid (PLA) material was used for component fabrication, chosen for its balance of lightweight properties and structural integrity, validated through stress analysis. A single limb’s forward and reverse kinematics were studied, enabling the development of optimized gait patterns. SIMSCAPE Multibody in MATLAB was utilized for dynamic simulations, and Proportional Derivative (PD) and Proportional Integral Derivative (PID) controllers were tested to evaluate trajectory tracking accuracy and stability. Results show that the six-legged configuration exhibits superior stability with a 15% improvement in gait cycle efficiency and a 20% reduction in energy consumption per stride compared to the four-legged counterpart. The use of PID controllers further enhanced performance, achieving a 12% improvement in settling time and reducing oscillations in trajectory tracking tasks. The choice of PLA material ensured durability under operational loads, with minimal deformation over repeated stress cycles. Servomotor selection and configuration were tailored to optimize torque and speed, enabling precise leg control. This study highlights the potential of bio-inspired robots to advance robotic mobility through optimized kinematics and material choices.

Enhancing Collision Avoidance in Mobile Robots Using YOLOv5: A Lightweight Approach for Unstructured Environments

Saleel H. Abood — 2025-04-09

Mobile robots play a crucial role in Industry 4.0, particularly in dynamic and unstructured environments where moving obstacles present significant challenges. This study applies the YOLOv5 object detection algorithm to enhance robotic perception and obstacle avoidance. The primary objective is to improve the accuracy and speed of object detection in real-time scenarios, ensuring safer and more efficient navigation for robots. The research contribution lies in developing a lightweight YOLOv5 model optimised for robotic applications, capable of detecting objects with high accuracy. The model was trained on a diverse dataset of 10,700 images, including static and dynamic objects such as chairs, fans, fire extinguishers, and humans, captured under various conditions and orientations. The dataset was divided into training (70%), validation (15%), and testing (15%) subsets. The proposed model achieved a mean average precision (mAP) of 0.73 at a confidence threshold of 0.374, demonstrating superior performance compared to the YOLOv4 model in terms of accuracy and processing speed. Notably, the model excelled in detecting static objects such as chairs, achieving a perfect recognition rate of 1.00, while encountering challenges with moving objects such as humans due to motion blur and rapid changes in body posture. These findings highlight the model’s potential for real-time applications in industrial and unstructured environments. In conclusion, this study demonstrates that the enhanced YOLOv5 model significantly improves object detection and collision avoidance capabilities in robotic systems.

Mobility Aid for the Visually Impaired Using Machine Learning and Spatial Audio

Wahyudi Wahyudi — 2025-04-09

Assistive technology is crucial in enhancing the quality of life for individuals with disabilities, including the visually impaired. Many mobility aids lack advanced features such as real-time machine learning-based object detection and spatial audio for environmental awareness. This research contributes to developing more intelligent and adaptable assistive technology for visually impaired individuals, promoting improved navigation and environmental awareness. This research presents a head-mounted mobility aid that integrates a time-of-flight camera, a web camera, and a touch sensor with K-Means clustering, Convolutional Neural Networks (CNNs), and concurrent programming on a Raspberry Pi 4B to detect and classify surrounding obstacles and objects. The system converts obstacle data into spatial audio, allowing users to perceive their surroundings through sound direction and intensity. Object recognition is activated via a touch sensor, providing distance and directional information relative to the user using audio description. The concurrent programming implementation improves execution time by 50.22% compared to Infinite Loop Design (ILD), enhancing real-time responsiveness. However, the system has limitations, including object recognition limited to 80 predefined categories, a 4-meter detection range, reduced accuracy under high-intensity sunlight, and potential interference in spatial audio perception due to external noise. Assistive technology to help the mobility of blind people using advanced technology based on machine learning has developed in a form that can be used flexibly for the user's mobility.

Extreme Learning Machine-Based Repetitive Proportional Derivative Controller for Robust Tracking and Disturbance Rejection in Rotational Systems

Enggar Banifa Pratiwi — 2025-04-10

Tracking periodic signals and rejecting periodic disturbances are common applications of repetitive control (RC). However, traditional RC methods struggle to compensate for aperiodic disturbances and adapt to system uncertainties, limiting their real-world effectiveness. Existing hybrid approaches often require extensive parameter tuning or suffer from high computational costs, creating a research gap in achieving both adaptability and efficiency. This paper proposes an improved control strategy called extreme learning machine repetitive proportional derivative control (ELMRPDC), which integrates repetitive proportional derivative control (RPDC) with an extreme learning machine (ELM). RPDC ensures accurate tracking of periodic signals, while ELM estimates and compensates for disturbances, enhancing overall performance. Unlike conventional neural network-based controllers, ELM enables rapid adaptation with minimal computational overhead, making it more suitable for real-time applications on resource-constrained systems. The proposed method is analyzed for stability using the Lyapunov approach, ensuring convergence of tracking errors. Extensive simulations are conducted on both rotational and linear dynamic systems under various disturbance conditions, including periodic, time-varying, multi-periodic, and aperiodic disturbances, such as vibration-induced disruptions in machinery. The study also evaluates the impact of hidden layer neuron variations in ELM on disturbance rejection. The best performance is observed for multi-period sinusoidal disturbances, achieving an RMSE of 1.8630 degrees at 1500 neurons, reducing error by 67.47% compared to conventional RPDC. These results highlight ELMRPDC’s advantages in computational efficiency, real-time feasibility, and robustness against complex disturbances. The approach holds significant promise for precise reference tracking and disturbance rejection across diverse industrial applications.

A Comparative Analysis of Numerical Techniques: Euler-Maclaurin vs. Runge-Kutta Methods

Mohammad W. Alomari — 2025-04-12

This study introduces a novel higher-order implicit correction method derived from the Euler-Maclaurin formula to enhance the approximation of initial value problems. The proposed method surpasses the Runge-Kutta approach in accuracy, stability, and convergence. An error bound is established to demonstrate its theoretical reliability. To validate its effectiveness, numerical experiments are conducted, showcasing its superior performance compared to conventional methods. The results consistently confirm that the proposed method outperforms the Runge-Kutta method across various practical applications.

A Software-Centric Evaluation of the VEINS Framework in Vehicular Ad-Hoc Networks

Jalal Mohammed Hachim Altmemi — 2025-04-12

The Large Communication Substitution between vehicles to infrastructure (V2I) or between vehicles (V2V), called as Vehicular Ad-Hoc Networks (VANETs), in new Tale Intelligent Transportation Systems (ITS) are degree and stop developing for extensive traffic manage, highway safety, and driverless cars. VEINS is a simulator framework that couples OMNeT++ and SUMO, widely used for assessing performance of VANET protocols and network architectures. However, it was observed that no existing research reviewed VEINS applications, limitations, or emerging trends in a structured manner. In this paper we provide a software-oriented summary of VEINS-based VANET studies. This research adds a taxonomy-oriented classification of studies published from 2011 to 2022, focusing on IEEE Xplore, ScienceDirect, and Scopus categorized security, safety, and other VANET applications. It identifies some gaps, right from the scalability, and computational overhead aspect to the limited integration of the next-generation technologies like 5G, Blockchain, and AI. A well-defined article selection strategy, extensive data extraction, and a comparative analysis of published VEINS studies is followed throughout the study. Statistical analyses show a growing percentage of VEINS but also point out obstacles to its real world usage. Key Insights point to an emphasis on security and safety, with little focus on emerging technologies and real-world validations. This review adds value to the body of knowledge by (1) establishing a systematic taxonomy of VEINS-based research, (2) highlighting gaps in research and methodological limits, and (3) providing future research directions focused on VEINS scalability enhancement, real-world validation frameworks, and AI-enabled VANET optimizations. The study is expected to be a useful reference for researchers and practitioners who intend to improve VEINS-based simulations of VANETs and accelerate development in the field of ITS.

Addressing Rogue Nodes and Trust Management: Leveraging Deep Learning-Enhanced Hybrid Trust to Optimize Wireless Sensor Networks Management

Santosh Anand — 2025-04-14

Comprising a multiplicity of AdHoc sensors working in concert to monitor a range of environmental and physical factors for the targeted area, wireless sensor networks (WSNs). These sensors are used to provide continuous environmental status like temperature, pressure, and humidity by forwarding vital data to the internet through a base station. Aiming to greatly increase the security and performance of WSNs, this study presents a new framework that is a combination of the Deep Learning-Enhanced Hybrid Trust (DLEHT) model and the Machine Learning-Enhanced Fuzzy-Based Routing Protocol (ML-EFBRP). In this research, enhanced packet delivery, packet drop reduction, and the rogue nodes addressed in WSN from source to sink using a probabilistic approach, which depends on the experience of data with the integration of a sum-rule weight mechanism in HMM (Hidden Markov Model). Integration methodology played a major role in deep learning to observe the normal and abnormal node behavior with historic data. It enhanced the throughput and lowered latency with successful detection and addressing of rogue nodes by the integrated strategy. The proposed work, reflects an improvement in performance, both in terms of throughput and latency. The delay hyperparameters are observed, which vary from 7.48 to 26.22 ms with an average of 15.855 ms. And the packet is controlled and decreased by 7%, showcasing more improvement compared to existing work. Simulation results show considerable improvements in network accuracy, reliability, energy efficiency, and resistance during node failures and security concerns for network correctness. These findings show the combination of DLEHT and ML-EFBRP models provides stronger monitoring systems, hence enhancing operational efficiency in settings with limited resources.

Optimal TID Tracking Control for Industrial Delta Robot Based on Harmony Search

Donia Saleem — 2025-04-16

This paper seeks to enhance the delta robot accuracy using the Tilt-Integral-Derivative (TID) technique. A CAD model for a real delta robot was developed on SolidWorks®. Then a Simscape model was generated using MATLAB® to apply the proposed control technique. The proposed TID technique was compared with the Proportional integral derivative (PID) control to ensure robustness. The harmony search (HS) optimization was used to find the optimal parameters of the PID and the TID controllers based on an effective objective function. Several operating points of robot angles were applied to investigate the accuracy of each control technique. The results show that the TID based on harmony search had the best settling time, rise time, minimum overshoot, and minimum steady-state error.

Design of a Novel Observer-Based SMC for WECS System Using PMSG to Obtain Maximum Energy

Dang Quoc Du — 2025-04-17

This paper studies and proposes a new sliding mode controller (SMC) for a wind energy conversion system (WECS) using a permanent magnet synchronous wind turbine generator (PMSG) to harvest maximum energy when the wind speed changes. In addition, the paper introduces a nonlinear disturbance observer (NDOB) to estimate the actual wind speed, to provide input to the proposed controller that the authors have studied. The control scheme proposed in this study not only considers the changes in the system parameters, but also considers the randomness of the wind speed changes. The effectiveness of the new SMC design and control scheme is demonstrated by simulation results on Matlab/Simulink software. These results are shown in the change of wind speed deformation, wind deformations that the turbine receives, turbulence assessment, the observer also estimated the non-snow parameters, the system also takes into account maximum power point tracking (MPPT), always consistent with the proposed control law. Moreover, the research results also show that the system works stably, the output is always close to the set value, the system works with high quality, thereby proving that the research results have been studied by the authors are suitable, bringing great benefits in the control process.

Integrating Multi-Sensors and AI to Develop Improved Surveillance Systems

Preeti Mohanty — 2025-04-19

This paper explores advancements in surveillance systems, focusing on the integration of multisensory and AI technologies in urban and environmental monitoring. It highlights the fusion of data sources such as video feeds, LiDAR, and wireless networks for enhanced real-time surveillance in complex environments. Artificial intelligence (AI) plays a critical role in anomaly detection, object identification, and behavior analysis, improving response times in high-traffic and security-sensitive areas. However, these technologies raise privacy concerns, emphasizing the need for responsible data management and ethical frameworks. Also, there is probability of false positives which can lead to unnecessary action disturbing the normal mode of life. These technologies involve high financial requirements hence must be used judiciously. In current study human surveillance is carried out in indoor environments by two AI algorithms: YOLOV5 and R-CNN. The results of these algorithms can be fused with LiDAR data for better decision making. R-CNN produced better results than YOLOV5 but the fusion with sensor data led to accurate detection of humans in indoor environments. R-CNN showcased better results than YOLOV5. The future of surveillance should focus on balancing safety and personal rights while adapting policies to ensure privacy and accountability in an increasingly tech-driven world.

Novel Hybrid SM Strategy Based on Speed Control and Disturbances Rejection for High Performance DSIM Drives

Ngoc Thuy Pham — 2025-04-22

In control theory and applications, disturbance cancellation is a critical challenge in the control of nonlinear drive systems, particularly in applications involving Dual Star Induction Motors (DSIM). This paper proposes a new adaptive hybrid sliding mode (SM) strategy that integrates a Repetitive Control (RC) scheme into an improved Second-Order Sliding Mode (SOSM) structure. The goal is to enhance tracking accuracy and periodic harmonic disturbance rejection in DSIM drive systems. The strategy also incorporates a load torque disturbance estimator that efficiently identifies and cancels disturbances, further improving system performance. System stability is guaranteed using Lyapunov theory, ensuring that the virtual control vectors for speed and current loops maintain stability throughout the operation. Simulation results using MATLAB confirm the effectiveness of the proposed control strategy, demonstrating improved tracking performance, harmonic disturbance rejection, and robust operation of the DSIM under varying conditions.

Effectual Energy Optimization, Fault-Tolerant Attack Detection, and Data Aggregation in Healthcare IoT Using Enhanced Waterwheel Archimedes and Deep Siamese Maxout Forward Harmonic Networks

Ganesh Srinivasa Shetty — 2025-04-25

The Internet of Medical Things (IoMT) has emerged as a transformative technology for improving healthcare delivery and patient outcomes. However, IoMT systems face significant challenges, including high latency, energy inefficiency, and vulnerability to cyberattacks, which compromise data security and patient privacy. Existing methods for attack detection and secure routing in IoMT often suffer from high latency, limited fault tolerance, and insufficient accuracy in identifying sophisticated attacks. To address these challenges, this paper proposes two novel approaches: the Improved Waterwheel Archimedes Optimization Algorithm (WWAOA) for secure routing and the Deep Siamese Maxout Forward Harmonic Network (DSMFHN) for attack detection in healthcare IoT. The Improved WWAOA integrates the Waterwheel Plant Algorithm (WWPA) with the Archimedes Optimization Algorithm (AOA) to optimize cluster head (CH) selection and secure routing. It considers key fitness parameters such as energy consumption, link lifetime (LLT), trust, delay, distance, and fault tolerance to enhance network efficiency and resilience. The DSMFHN combines Siamese Neural Networks (SNN) and Deep Maxout Networks (DMN) with forward harmonic analysis to detect attacks with high accuracy and low false positive rates. Additionally, data aggregation is performed using Bidirectional Long Short-Term Memory (BiLSTM) with adaptive weightage based on fault and malicious node detection. Experimental results demonstrate that the proposed methods outperform existing techniques. The Improved WWAOA achieves a minimal delay of 0.557 ms, maximal energy efficiency of 0.182 J, a packet delivery ratio (PDR) of 93.894%, and a trust value of 87.152. Meanwhile, the DSMFHN achieves a high accuracy of 92.598%, a true positive rate (TPR) of 91.643%, and a low false positive rate (FPR) of 0.156. These results highlight the effectiveness of the proposed methods in addressing the critical challenges of latency, energy efficiency, and security in healthcare IoT systems.

Disturbance Handling and Efficiency Optimization for SPWM-Three Phase Inverter by Using PID Controller System

Suaad Makki Jiaad — 2025-04-27

The importance of studying inverters in electrical systems is highlighted by their role as one of the most important electronic power devices used in numerous applications in industry, as well as in generation, transmission, and distribution, most notably in renewable energy generation systems. An inverter converts direct current power into alternating current power to power loads or connect solar energy sources to the grid. Inverters are built using electronic switches such as thyristors or transistors such as IGPTs and MOSFET transistors. A number of switches are used to build the inverter, depending on the type of inverter, whether single-phase or three-phase. It can also be half-wave or full-wave. The current study proposed a bridge inverter consisting of six electronic switches of the IGBT transistor type arranged in two rows and three columns. To operate the inverter, pulse width modulation (PWM) technology was used to regulate the inverter operation and obtain the required output to supply a three-phase resistive load. In addition, an LC filter was connected to obtain a pure sine wave. Due to the different and variable operating conditions and to overcome disturbances, a conventional control unit was added to improve performance and raise the efficiency of the system. After conducting the proposed tests, the possibility of obtaining an inverter that operates with an efficient system to cover the load requirements under variable operating conditions was verified.

Integration of Sparrow Search Optimization with Terminal Synergetic Control for Permanent Magnet Linear Synchronous Motors

Mohanad Nawfal — 2025-04-29

This paper proposes a theoretical framework of the procedure to design an optimal robust terminal synergetic control (TSC) for the permanent magnet linear synchronous motors (PMLSMs). The general component and the mathematical equations of the PMLSM are first introduced. Based on the established model of the PMLSM, the control law of the TSC is developed. The tuning process of the TSC gains is enhanced by employing sparrow search optimization (SSO) based on the Integral Time of Absolute Errors (ITAE). The effectiveness of the proposed control algorithm has been veriﬁed by numerical simulations using MATLAB software for a step input. Additionally, the results have been compared with the classical synergetic control (CSC). The comparison shows that the TSC exhibits a good performance in normal operation and in a robustness test involving system parameters’ changes as compared to the CSC.