Journal of Electronic Research and Application

Research on Multi-Rotor UAVs in Complex Indoor Environments

2026-02-13T09:38:14+08:00

The exponential growth of unmanned aerial vehicle (UAV) technology has spurred its adoption in diverse indoor applications, including infrastructure inspection, automated logistics, and emergency response. However, navigating through indoor environments, characterized by static obstacles, dynamic interferences, and spatial constraints, poses significant challenges to path planning algorithms. Developing efficient, robust, and real-time path planning solutions is crucial for enabling reliable autonomous UAV operations in such complex scenarios. This study presents a systematic approach to indoor UAV navigation, integrating custom hardware development, algorithmic innovation, and multi-faceted validation. An indoor UAV experimental platform was constructed around the Pixhawk 2.4.8 flight controller, complemented by a Firefly Core-3588L onboard computer, PXYZ-D435 depth camera, and OptiTrack motion capture system. After rigorous PID tuning and endurance testing, stable autonomous flight control was achieved via the Robot Operating System. Subsequent real-world tests on the custom UAV platform, involving obstacle courses and narrow passage traversals, further confirmed its robustness and stability in complex indoor environments. Overall, this research provides a practical framework for enhancing UAV navigation capabilities, with direct implications for real-world applications in logistics, surveillance, and emergency response.

NI-HotStuff: A Reputation-Driven Committee Framework for Efficient and Robust BFT Consensus

2026-02-13T09:38:48+08:00

Consensus mechanisms are fundamental to blockchain systems, ensuring that distributed nodes agree on the validity of transactions and data. However, performance bottlenecks, particularly those related to throughput, latency, and node selection, have increasingly constrained the scalability of modern blockchain deployments. To address these issues, this paper proposes NI-HotStuff, a reputation-driven committee-based BFT consensus framework built upon the HotStuff protocol. A CatBoost-based reputation model is introduced to learn and evaluate historical behavioral features of nodes, enabling quantitative reputation scoring. A hardware-aware bidding mechanism is further incorporated to dynamically compute each node’s bid value and integrate it with its reputation score, thereby prioritizing stable and high-performance nodes for consensus participation. Moreover, a committee mechanism is established in which a set of committee nodes were selected from the candidate pool, and only committee members participate in the consensus process, reducing redundant communication and mitigating the performance drag caused by weak nodes. On top of that, a leader-selection strategy based on reputation values and inter-view time intervals is designed to prevent low-reputation or potentially malicious nodes from frequently becoming leaders. Experimental results demonstrate that NI-HotStuff significantly outperforms traditional PBFT and HotStuff in terms of communication overhead, consensus latency, and system throughput, with particularly notable improvements in small- and medium-scale node environments.

A Small-Sample Bearing Fault Diagnosis Method Based on Multi-Image Fusion and Multi-Scale Dynamic Residual Dual Attention Mechanism

2026-02-13T09:38:45+08:00

In recent years, fault diagnosis methods based on convolutional neural networks (CNNs) have garnered significant attention in the field of rotating bearing fault diagnosis. Addressing the challenge of extremely limited fault signal samples, this paper proposes a small-sample bearing fault diagnosis method based on multi-image fusion and a dual-attention mechanism incorporating multi-scale dynamic residuals. This method first converts the fault signal into a two-dimensional image through continuous wavelet transform and Gram angle field (GASF/GADF), thereby transforming the fault diagnosis problem into an image feature learning problem. The model extracts basic features through the initial convolutional layer and sequentially learns deep features via multi-scale dynamic residual blocks and dual attention mechanisms. Among these, the multi-scale architecture captures features across different receptive fields through parallel convolutional branches, while the dual attention mechanism performs feature recalibration in both the channel and spatial dimensions. Experimental results demonstrate that the proposed method achieves an accuracy rate of 97.47% in bearing fault diagnosis tasks, representing a significant improvement over traditional CNN models. This validates the model’s effectiveness and superiority in complex fault diagnosis scenarios.

Overview of the Integration of Large Language Models, Knowledge Graphs, and GraphRAG, along with Research on Their Industrial Applications

2026-02-13T09:38:08+08:00

In recent years, Large Language Models (LLMs) have rapidly advanced in language understanding, reasoning, and generation, and are increasingly adopted as the “brain” of industrial intelligent systems. Nevertheless, in high‑risk and strongly regulated domains they still exhibit hallucination, weak domain grounding, limited interpretability, and privacy as well as security constraints. Knowledge graphs (KGs) encode domain entities, relations, rules, and events explicitly, providing controllable semantics and an explainable reasoning substrate. Retrieval‑augmented generation (RAG) injects external evidence into LLM prompting, while GraphRAG further introduces graph indexing and community‑level retrieval to preserve global structure and support multi‑hop reasoning. This review summarizes the evolution of LLMs, KG modeling and extraction, GraphRAG mechanisms, and a general fusion framework. Typical industrial applications are surveyed, and a coal mine flood emergency plan generation and evaluation approach is discussed to illustrate the practical value of graph‑grounded large models. KG‑enhanced retrieval also supports provenance tracking, allowing industrial users to audit the evidence behind model outputs.

Collaborative Quality Management in Industrial Engineering from a Supply Chain Perspective: AIDriven Enterprise Quality Optimization

2026-02-13T09:38:43+08:00

Amidst the intensifying digital economy and global competition, supply chain quality management is evolving from traditional linear models toward networked systems characterized by data-driven and intelligent collaboration. This paper constructs an AI-driven “Supply Chain Quality Collaborative Management” framework through system optimization and artificial intelligence analytical capabilities from a supply chain perspective. The study first analyzes core challenges in supply chain quality collaboration across three dimensions: data fragmentation, standard discrepancies, and mechanism asymmetry. It highlights that traditional static and reactive quality controls struggle to adapt to complex, dynamic supply chain ecosystems. Subsequently, through systematic literature review and theoretical synthesis, the paper elucidates AI’s role in multi-source quality data fusion, semantic alignment, standardized governance, and intelligent incentives. It proposes collaborative optimization pathways based on deep learning, blockchain, and reinforcement learning. Through case studies in the automotive and pharmaceutical industries, the research validates the feasibility of AI in predictive maintenance and cross-linkage collaborative decision-making, demonstrating AI’s ability to significantly enhance the systemic resilience and decision-response capabilities of quality management. This paper innovatively integrates industrial engineering process optimization with cross-organizational governance mechanisms for supply chain quality management, providing a new theoretical framework and practical pathway for intelligent manufacturing and sustainable supply chain development.

Application of the SCSSA-VMD Denoising Method in Natural Gas Pipeline Leakage Detection

2026-02-13T09:38:17+08:00

The decomposition performance of variational mode decomposition (VMD) on natural gas pipeline leakage pressure signals is highly sensitive to the subjective selection of its key parameters: the number of modes K and the penalty factor α. To address this issue, this paper proposes an enhanced sparrow search algorithm (SSA) that integrates sine/cosine searching and Cauchy mutation strategies, referred to as SCSSA, for optimizing the VMD parameter combination. Experimental results demonstrate that the SCSSA-optimized VMD method significantly outperforms denoising approaches based on the standard SSA and particle swarm optimization (PSO) in optimizing VMD parameters. Specifically, the proposed method achieves a higher signal-to-noise ratio (SNR) and a lower root mean square error (RMSE) in the denoised signal, effectively enhancing the denoising performance.

A Framework for Advancing Intelligent Electrical Agricultural Machinery Technologies in Mountainous Regions

2026-02-13T09:38:40+08:00

The complex terrain of the hilly and mountainous regions in southwestern China presents significant challenges to agricultural mechanization, resulting in a level that is markedly lower than the national average. Focusing on the key development needs for intelligent agricultural machinery in these areas. This paper systematically delineates four core technological domains: lightweight machine design, detachment and drag reduction in heavy clay soils, slope adaptability, and remote operation and maintenance. The study aims to provide technical insights for overcoming the bottlenecks in mechanizing hilly and mountainous agriculture, thereby contributing to rural revitalization and national agricultural development strategies.

Intelligent Inspection and Closed-loop Management Innovation of Campus Fire Safety

2026-02-13T09:38:56+08:00

As universities expand in scale and diversify in functions, traditional fire safety management faces challenges such as inefficient inspections and difficulty in tracing hazards. Technological innovation is key to enhancing efficiency, focusing on the core needs of campus fire safety. Intelligent inspection technology is applied to facility monitoring, hazard identification, and risk alerts, establishing a closed-loop system of monitoring, identification, disposal, and feedback. By leveraging the characteristics of IoT, AI, and big data technologies, the innovative solutions encompass technological integration, process optimization, and responsibility enhancement. This research aims to help universities overcome traditional bottlenecks, strengthen fire safety safeguards, and drive management transformation toward precision, intelligence, and efficiency.

Innovative Research on Construction Models and Operational Management of Electrical and Electronic Laboratories

2026-02-13T09:38:38+08:00

Electrical and electronic laboratories are crucial for developing engineering talent, yet they face challenges such as outdated hardware, rigid management, and faculty shortages. This paper proposes an integrated reform model featuring virtual-physical equipment upgrades, open and intelligent management platforms, a dual-qualified teaching team, and a full-process safety assurance system. It offers a practical framework for modernizing such laboratories and supporting the cultivation of high-quality innovative engineering professionals.

Design of Integrated Brushless Motor Drive and Control System for Robotic Arm Joints

2026-02-13T09:38:53+08:00

To meet the requirements of high performance, low cost, and modularity in robotic arm joints, this study designs an integrated brushless motor drive and control system. The system selects the STM32G473 microcontroller as the control chip and adopts field-oriented control as the primary motor control algorithm. Meanwhile, the drive and control system design is completed from both hardware and software aspects. Finally, the study performs closed-loop experiments on the robotic arm joint. The experimental results demonstrate that the designed drive and control system for robotic arm joints exhibits a favorable dynamic response and steady-state performance, making it suitable for controlling desktop-level and lightweight robotic arms.

Research on Motion Simulation of Panda Manipulator Based on ROS2

2026-02-13T09:38:50+08:00

Focusing on the research issues of path optimization and collision avoidance in robotic arm motion planning, this paper will construct a high-fidelity simulation environment using the ROS2 framework. By integrating MoveIt, Rviz visualization tool packages, and the Panda robotic arm URDF file, along with leveraging the DDS communication mechanism of ROS2, low-latency data interaction between the planning module and the simulation environment is achieved. The upper computer software is developed to conduct simulation studies on the path planning and trajectory interpolation principles of the robotic arm, thereby verifying the reliability of the ROS2 distributed architecture in robotic arm simulation.

Technology of Radar Detection

2026-02-13T09:38:35+08:00

Radar detection technology utilizes radio waves for target detection, localization, and identification. It involves emitting electromagnetic waves and receiving the reflected echoes from targets, then analyzing the echo characteristics to obtain target information. This paper focuses on the fundamental principles, advantages, and disadvantages of radar detection technology. It emphasizes synthetic aperture radar (SAR), passive radar detection technology seeker, and millimeter-wave active homing guidance target identification technology, as well as the characteristics and development status of other detection methods such as phased array radar, showcasing the multi-directional development trend of radar detection technology.

Design and Implementation of Machine Learningbased Monitoring System for Mineral Processing Flotation Reagent

2026-02-13T09:38:31+08:00

Flotation, also known as froth flotation, is a method for separating minerals from powdered materials by altering their floatability through the use of flotation reagents. This paper proposes a flotation process control system for mineral processing based on machine learning. Addressing the issue of lack of precise detection methods in the flotation process of iron concentrate, a neural network regression method is used to predict the amount of reagents and the grade of the flotation concentrate. The flotation data in this paper come from the Key Laboratory of Multitechnology Resource Utilization of Bayan Obo Mine, Inner Mongolia Autonomous Region. The preprocessed data form the dataset used to create the production prediction model. The neural network model is constructed using the PyTorch deep learning framework. Finally, based on the established model, a comprehensive flotation dosing monitoring system is developed using the Django framework, which includes functions such as production indicator large screens, workshop personnel safety monitoring large screens, flotation reagent usage processing, flotation reagent procurement platform.

Exploration and Practice in the Construction of Electrical and Electronic Laboratories

2026-02-13T09:38:33+08:00

This paper delves into the development of electrical and electronic laboratories in higher education institutions, elucidating their significance for talent cultivation and disciplinary advancement. It analyzes current challenges such as outdated equipment, imperfect practical teaching systems, and backward laboratory management models. A series of targeted strategies are proposed, including advancing equipment modernization, improving practical teaching systems, strengthening faculty development, and establishing open-access platforms. Implementation steps and effectiveness evaluation methods are introduced to create high-quality electrical and electronic laboratories that meet new-era demands, thereby enhancing university teaching, learning, and research standards.

Nasopharyngeal Carcinoma Lesion Recognition Based on Multi-Window Resampling Technology

2026-02-13T09:38:29+08:00

Accurate deep learning-based detection of nasopharyngeal carcinoma (NPC) magnetic resonance (MR) images is conducive to diagnosis and treatment. These images are characterized by high dimensionality, complex noise interference, and blurred tissue structure boundaries. How to extract key pathological features from massive imaging information and provide quantitative basis for clinical diagnosis remains an important challenge in the current field of medical image processing. This paper uses multi-window fusion technology to map multiple key window information to the pseudo-color space, realizing the integration of multi-dimensional feature information and compensating for the information limitations of single-window imaging. Experiments show that this method can effectively improve model accuracy.

Construction of a Motor-Driven Experimental Platform for Exploring the Law of Light Reflection

2026-02-13T09:38:27+08:00

In traditional middle school optical experiments, the fixed light source with an iron stand is inconvenient to operate, and the water mist generated by a spray bottle has a short duration and easily affects the reflection effect, leading to many limitations in the experimental exploration of the law of light reflection. This paper constructs a motor-driven experimental platform for exploring the law of light reflection. It innovatively adopts motor drive to realize flexible adjustment of the light source angle, and uses a medical humidifier atomizer instead of a traditional spray bottle to ensure continuous and stable mist that is not easy to adhere to the mirror surface. Through modular design, the platform integrates the functions of light source adjustment, mist generation and reflection observation. It has a simple structure and convenient operation, effectively solving the pain points of traditional experimental devices, providing a more efficient practical tool for optical experiment teaching, and featuring low cost and easy promotion.

Exploring the Path of AI Technology’s Empowerment of New Developments in Higher Education

2026-02-13T09:38:25+08:00

As the core driving force leading the new round of technological revolution and industrial transformation, artificial intelligence is profoundly reshaping the higher education ecosystem. Based on the background of artificial intelligence development, this paper expounds the value of AI technology in promoting the innovative development of higher education, explores the specific paths of AI technology empowering the innovative development of higher education from three dimensions of talent training, scientific research, and governance, and puts forward the required guarantee conditions. It aims to promote the in-depth integration of artificial intelligence and higher education, profoundly change the form of higher education, lead higher education towards a more personalized, precise, and intelligent development path, and provide reference for solving a series of problems encountered in the current development of higher education.

RoboMirror: Bridging Human Intent and Robot Capability Through Self-Reflective Motion Adaptation

2026-02-13T09:39:56+08:00

Current humanoid robot control paradigms place the burden of feasibility assessment on human operators, who must carefully design commands within perceived robot limitations. This constraint significantly hinders practical deployment and limits the expressiveness of robot behaviors. This study proposed an inverting paradigm: rather than constraining operator inputs, robots should autonomously evaluate their capacity to execute commanded motions and intelligently adapt references to align with their physical constraints and learned skills. This study introduced the Performance Prediction Network (PPN), a transformer-based architecture that forecasts execution quality for arbitrary reference trajectories by analyzing both the commanded motion sequence and current robot state. Given a high-level task specification, our framework synthesizes multiple viable motion candidates and employs PPN to rank them across six dimensions: collision avoidance, kinematic feasibility, dynamic stability, trajectory smoothness, and goal satisfaction. This ranking enables autonomous selection of the most suitable reference motion before execution begins. Our complete system integrates motion generation, kinematic retargeting, and learned control policies with PPN-guided adaptation, creating a closed-loop framework where robots reason about their own limitations. Validated on 100,000 diverse human motions span walking, running, jumping, and acrobatic maneuvers, PPN achieves 99.14% accuracy in predicting imminent failures while maintaining low prediction error across all performance metrics. In deployment, our system successfully prevents 62% of anticipated falls by autonomously modifying commanded references, demonstrating that explicit capability modeling enables safer and more reliable humanoid control without sacrificing behavioral diversity.

Transmission Line Defect Detection Algorithm Based on Improved RT-DETR Model

2026-02-13T09:38:10+08:00

This paper addresses the urgent need for high-precision and high-efficiency visual perception technologies in power equipment operation and maintenance under the background of rapid development of smart grids. It points out the performance limitations of the existing real-time target detection framework RT-DETR when handling small targets, dense targets, and complex backgrounds in power inspection scenarios. To overcome this bottleneck, this study proposes an improved backbone network model, DETR-EVA, based on an efficient visual attention mechanism (EVA). This model innovatively designs an attention computation structure with linear complexity by deeply integrating the EVA mechanism with the C2f module in the RT-DETR backbone network, and combines local detail perception and global dependency modeling capabilities. Its core lies in the introduction of a gated fusion mechanism, which significantly enhances the model’s ability to model long-distance contextual relationships and the adaptive adjustment efficiency of feature weights while retaining the advantages of multi-branch feature extraction and fusion of the C2f module. Experiments were conducted on an inspection image dataset containing typical power equipment targets. The results show that compared with the original RT-DETR model, DETR-EVA improves the overall accuracy index mAP50-95 by 2.5%, reduces computational complexity by 14%, and reduces the number of model parameters by 27%. This effectively verifies that the proposed method can significantly improve the detection accuracy of small targets and complex scenes while maintaining real-time detection speed, providing a better visual solution for intelligent operation and maintenance of power equipment.

Deep Learning-Based Highway Rockfall Early Warning System

2026-02-13T09:38:21+08:00

This paper proposes a deep learning-based rockfall warning system for mountainous road curves. It utilizes drone inspections combined with the YOLOv11 object detection algorithm to accurately identify rockfalls on road surfaces, while employing ground-based millimeter-wave radar for real-time vehicle detection. The system features a comprehensive curve blind spot warning mechanism and incorporates a wireless communication module to push instant alerts to mobile navigation terminals based on rockfall risk and vehicle location. This system effectively addresses the challenges of rockfall identification and delayed warnings within blind spots on curves. It reduces manual inspection costs while significantly enhancing driving safety on mountainous roads.

Research on Unmanned and Intelligent Combat Theory and Capability Development

2026-02-13T09:38:19+08:00

This paper presents a comprehensive analysis of the evolution, foundational concepts, capability development, and operational challenges of unmanned systems. It traces their theoretical progression from post-Cold War origins to systematic maturation in the 21st century, emphasizing the central role of emerging operational concepts and exploratory advances in capability development. Key bottlenecks in unmanned combat operations are examined, particularly limitations in communication bandwidth and the vulnerability of data links in contested environments. The paper further discusses future development trajectories, highlighting both technological and ethical challenges. Overall, unmanned warfare is evolving toward a more intelligent, networked, and resilient operational architecture, with profound implications for the conduct and character of future high-end warfare.

Application Research of Concept Bottleneck Model in Passport Printing Method Detection

2026-02-13T09:38:12+08:00

With the increase in cross-border mobility, passports, as critical identity documents, require robust anti-counterfeiting security. While existing deep learning-based automatic detection methods achieve high accuracy, they lack interpretability. This paper introduces the Concept Bottleneck Model (CBM) to construct a transparent passport printing method detection framework. By defining interpretable intermediate concepts and integrating linear reasoning, the model significantly enhances reliability and debugging efficiency. The article systematically analyzes the advantages, challenges, and future directions of this approach.

Design and Simulation of a Microstrip Frequency-Scanning Antenna for Millimetre‑Wave Fuze Applications

2026-02-13T09:38:15+08:00

To satisfy the simultaneous requirements of high gain and wide angular coverage for millimetre‑wave fuzes under large impact‑angle variations, this paper proposes a microstrip frequency‑scanning antenna based on a quasi‑travelling‑wave, series‑fed patch array. The antenna is implemented on Rogers 4350B substrate (εr = 3.5, thickness h = 0.254 mm) and operates over 30–36 GHz. By exploiting the frequency‑dependent phase progression along the series feed, the main beam steers continuously without phase shifters. Full‑wave simulations in HFSS show that the antenna maintains |S11| < −10 dB across the entire band. The E‑plane main beam scans from 48° at 30 GHz to 0° at 36 GHz, providing a 48° frequency‑scanning range; when the 3‑dB beamwidth is included, the effective detection‑angle coverage reaches approximately 74°. The simulated gain remains stable above 10.5 dBi, peaking at about 11.6 dBi near 35 GHz. With a low‑profile, planar structure (overall size ≈ 20 mm × 10 mm) and no additional terminal load, the proposed design offers a compact solution for fuze antennas that require broad angular coverage and robust gain in the 30–36GHz.