IJMEE Journal Cover

International Journal of Mechanical and Electrical Engineering

ISSN: 3005-9615 (Print)ISSN: 3005-7132 (Online) DOI: 10.62051/ijmee Frequency: Monthly

International Journal of Mechanical and Electrical Engineering (IJMEE), an English‑language open‑access journal published by Warwick Evans Publishing, provides a rigorous peer‑reviewed platform for the dissemination of original work - whether empirical or theoretical - that contributes to knowledge in mechanical engineering, electrical engineering, and their interdisciplinary fields. The journal spans the entire scope of these disciplines.

Scope: Acoustics, aerodynamics, electronics, manufacturing, control & monitoring, fracture‑fatigue mechanics, fluid‑thermal engineering, mechatronics, robotics, signal processing, structural mechanics, vehicle technology, etc.

Indexing & Abstracting: Harvard Library, Crossref, ResearchGate, Scilit, Google, Mendeley, Semantic Scholar, etc.

Latest Articles

Uncertainty-Aware Congestion Reward Shaping for Multi-Objective IC Placement with Deep Reinforcement Learning

Abstract: Reinforcement-learning (RL) placers increasingly use learned congestion predictors to densify their reward through potential-based shaping. Existing congestion-aware shaping evaluates an overflow penalty at a point prediction of the final routing demand, although this demand is uncertain while objects remain unplaced and the predictor is unreliable on unfamiliar states. We show that the plug-in potential is optimistically biased by Jensen’s inequality, most strongly early in an episode, and propose UA-CARS: a deep ensemble of heteroscedastic grid-graph GCNs predicts the final g-cell utilisation, the expected overflow penalty is evaluated in closed form, and the potential is shrunk toward a reference potential with a precision weight derived from ensemble disagreement, which preserves policy invariance and adds no tuning parameter. On four OpenROAD designs built from public gate-level netlists, the uncertainty-aware potentials reduce the error of the congestion-to-go estimate by 32–36% and lower their trust automatically where the predictor generalizes poorly. A controlled PPO study with 56 training runs shows, however, that potential accuracy alone does not yield better placements: with a neutral reference, UA-CARS performs on par with the plug-in baseline, and both are outperformed by the exact incremental congestion reward. Anchoring the shrinkage on that reward (UA-CARS-M) matches it and lowers the late-training cost of CARS by 5.6% (11/12 design–seed pairs, p=0.002). We trace these results to the accuracy of potential differences, which governs the benefit of shaping. Read More

Task-Adaptive Dual-Graph Neural Network for Joint Congestion and Timing Prediction

Abstract: Early physical-design decisions benefit from jointly estimating routing demand and timing risk, but a netlist edge and a geometric-neighbor edge encode different phenomena. We present TD-DGNN, a task-adaptive dual-graph predictor with separate spatial and netlist GraphSAGE branches, learned congestion- and timing-specific fusion gates, and a directed congestion-conditioned timing head. To isolate architectural effects, we construct 144 independently generated synthetic placed directed-acyclic circuits (18,311 cells), using a rectangular uniform wire-density estimator and a congestion-detour longest-path timing engine. Nine configurations, including a capacity-matched baseline, are trained with three seeds and assessed on 24 unseen circuits. TD-DGNN attains mean per-design timing correlation 0.855 and violation ROC-AUC 0.825 versus 0.814 and 0.810 for a parameter-matched single-graph reference; the paired timing gain is 0.041 (95% bootstrap interval 0.032–0.050). Without retraining, on 24 larger synthetic circuits it attains correlation 0.836 versus 0.797 for that reference. Directional predecessor aggregation adds 0.0063 timing correlation relative to undirected coupling. The independent effect of adaptive gates and congestion-conditioning is smaller on this benchmark. These results establish a controlled test of graph separation rather than an estimate of commercial sign-off accuracy. Read More

Model-Based Quantized Edge Vision Inference: Component Detection System for the STM32N6 Circuit Diagram

Abstract: Against the backdrop of the rapid development of Edge AI, deploying object detection models on resource-constrained microcontrollers (MCUs) presents simultaneous challenges in terms of computing power, memory, and accuracy. This paper proposes an improved algorithm for detecting components in circuit diagrams and deploys it on an STM32N6 edge device. At the algorithmic level, using YOLOv8n as the baseline model, the study addresses the characteristics of circuit diagram components—which are extremely small and often arranged horizontally or vertically—by introducing an NWD-SIoU joint loss function to replace the default CIoU loss, thereby providing stable gradient signals and directional alignment for tiny objects.Additionally, Dropout regularization is incorporated into the detection head structure of YOLOv8n to address overfitting in circuit diagram training scenarios. Compared to the original algorithm, the improved algorithm achieves a 0.036 increase in F1-score, with precision and recall improving by 0.017 and 0.031, respectively. At the deployment level, post-training quantization from FP32 to INT8 was completed, reducing the weight size from 12 MB to 3.23 MB. To address the limited SRAM capacity of the STM32N6, a heterogeneous storage hierarchy allocation scheme was designed: static weights are stored in external Flash, while runtime activation tensors are locked in the NPU’s dedicated RAM. This approach frees up SRAM capacity while accelerating model execution. Experimental results demonstrate that the proposed approach achieves a balance of size, accuracy, and speed on the STM32N6 platform. It validates the feasibility of running complex vision models on edge devices and provides a reusable technical path for the implementation of edge AI in the field of industrial vision inspection. Read More

Evidence-Calibrated Hierarchical Repair of Combinational RTL under Bounded Verification Budgets

Abstract: Verification-guided repair prioritizes candidate edits using evidence from failed executions, but its effectiveness under bounded verification budgets depends on both the fault model and the search schedule. We investigate evidence-calibrated hierarchical repair (ECHR), combining backward output-cone localization, four implemented semantic counterexample tags, operator-family priors, feedback-based reranking, guarded two-edit branches, and exact finite-domain acceptance. The tag classifier, operator priors and cone localization adapt the counterexample-signature machinery of the VGMA verification-guided multi-agent RTL framework. Across ten deterministic synthetic-corpus seeds, we execute 2,520 detectable faults on 180 small combinational reference circuits, including 1,080 two-site faults. Every candidate is evaluated on 64 witness vectors, and accepted repairs must agree with the reference for all 512 three-input, three-bit assignments. With a fixed 30-query budget, ECHR repairs 820/2,520 cases (32.54%), compared with 400/2,520 (15.87%) for random search, 762/2,520 (30.24%) for static combined ranking and 771/2,520 (30.60%) for the strongest static ranking (tag-only). The ECHR–combined-ranking difference is positive in all ten seeded replications (mean +2.30 percentage points; 95% seed-bootstrap interval +1.67 to +2.94). Ablations show complementary effects of feedback reranking and two-edit expansion, although two-site repair remains limited to 53/1,080 cases. Results characterize a restricted, synthetic bit-vector repair problem rather than sequential or industrial Verilog. Read More

Coverage-Balanced Field-Intensity Sampling for Physics-Informed Electrostatic Simulation

Abstract: Physics-informed neural networks (PINNs) can under-resolve electromagnetic domains when physics residuals and predicted field variations are used to greedily rank collocation points. We instantiate a published geometry-initialized, residual/field-gradient sampling procedure in a two-layer coaxial electrostatic benchmark and investigate two corrective choices: score the electric-field intensity rather than its spatial derivative, and impose equal-area coverage quotas with uniform exploration. The resulting coverage-balanced mixed sampler (CBMS) solves two Cartesian-coordinate Laplace subproblems with dielectric interface constraints; closed-form electromagnetic fields are held out exclusively for testing. Across three geometric/material regimes, four primary policies are evaluated using three paired random seeds and 3,600 optimizer updates. In the narrow-inner-layer regime, field-gradient-based global ranking and its coverage-balanced counterpart are unstable, whereas intensity-CBMS yields an error of 0.094 compared with 1.030 for the gradient-based global comparator and 0.050 for uniform sampling. The distinct material and geometry configurations reveal that no strategy dominates every regime. Multi-seed sampling ablations, field-error diagnostics, and exact synthetic reference tables document both the gains and the limits of coverage-aware adaptivity. Read More

Research on the Development of an AI Literacy Curriculum Framework for Maritime-Oriented ICT Majors

Abstract: This paper addresses the insufficient alignment between AI literacy education and disciplinary learning, the limited integration of industry-specific scenarios, and the overemphasis on tool operation in assessment at vocational colleges with distinctive industry orientations. It develops a “1+M+N” curriculum framework guided by AI4M (Artificial Intelligence for Marine). The framework establishes a shared foundation across related majors through the Introduction to Artificial Intelligence course, strengthens professional skills through modular courses, and promotes the transfer of learning through practical maritime projects. The AI literacy course comprises four instructional modules, with ethics education integrated throughout. Its implementation combines college–industry collaboration, scenario-driven learning, assessment of student work, and feedback on learning progress. The resulting curriculum design integrates general AI literacy, professional ICT (information and communication technology) competencies, and responsibility in maritime practice, providing a reference for AI literacy education at industry-oriented vocational colleges. Read More

Improved SVPWM for Torque Ripple and Harmonic Suppression of PMSM Drives

Abstract: To address the inherent torque ripples and harmonic distortions existing in conventional permanent magnet synchronous motor (PMSM) drive systems, this paper proposes an improved space vector pulse width modulation (SVPWM) strategy for performance optimization. Different from the traditional hysteresis-based control scheme, the proposed method adopts proportional-integral (PI) regulators to achieve stable and accurate closed-loop control. The operating principle of the SVPWM algorithm is elaborated through mathematical derivation, and optimal reference voltage vector are selected to refine the output voltage characteristics. A complete PMSM drive simulation model is established to validate the dynamic and steady-state performance. The results show that the improved SVPWM strategy achieves excellent torque stability under both no-load and loaded conditions, where the torque fluctuation is limited within 1 N·m. The motor can rapidly track the rated speed during startup and maintain stable operation under sudden load variations. Meanwhile, the stator current presents a smooth waveform with minor peak fluctuations and no obvious distortion. Simulation results validate that the improved SVPWM strategy can effectively suppress torque ripples and reduce current harmonic distortion, while enhance the dynamic response and anti-interference capability of PMSM drive system. Read More

Design of a Offshore Floating Waste Recovery Device based on FBS-TRIZ

Abstract: In response to the problems of low recycling efficiency and insufficient environmental adaptability of the floating garbage collection equipment in the near sea area, the FBS-TRIZ integrated method was adopted for the design. Through on-site research, user demands were obtained and screened. The FBS model was used to complete the layer-by-layer mapping of demands to functions, behaviors and structures. Based on the mapping results, technical and physical contradictions were identified, and solutions were sought using the TRIZ contradiction matrix and inventive principles. On this basis, design schemes such as a double hull body, adjustable garbage collection mechanism, inclined slope type conveying and dehydrating mechanism, and intelligent control module were formed, achieving the synergy of goals such as garbage collection, stable navigation, low energy consumption and ecological protection. The results show that this method can provide a systematic design path for the conceptual design and structural optimization of the floating garbage collection equipment in the near sea area. Read More

An Integrated Navigation and Obstacle Avoidance Scheme for Indoor Mobile Robots

Abstract: Indoor mobile robots are expected to reach a goal, build a usable representation of the surrounding space, and avoid obstacles that were never present on any prior map — all simultaneously, and on hardware compact enough to be carried on board the robot. Most published navigation stacks rely on a desktop computer or an embedded Linux board running ROS, which is convenient for research but heavy, power-hungry, and difficult to certify for real-time behavior. This report describes a navigation and obstacle-avoidance scheme that runs in its entirety on a single STM32F407 microcontroller, with no operating system underneath it. The control logic was built using a model-based design (MBD) flow: peripherals were configured in STM32CubeMX, the application layer was modeled in MATLAB/Simulink, and embedded C was generated automatically and flashed to the board. On this platform we layer an extended Kalman filter for pose estimation, an occupancy-grid representation refreshed from a 2-D LiDAR, a grid-based A* planner for the global route, and a dynamic-window (DWA) reactive layer that handles conditions the global map did not anticipate. Switching between plan-following and obstacle avoidance is handled by a supervisory state machine. We ran 150 trials across five indoor scenarios. On average the scheme reached the goal in 92.7% of runs, held localization error under about 0.08 m around a 38 m loop, and still kept a 28% timing margin inside the 20 ms control cycle. Thanks to conditional compilation, the same model runs untouched both in offline simulation and on the physical chip — and this capability, more than any individual algorithm, was central to keeping development tractable. Read More

Experimental Strain Correction of a Cantilever Beam Using DQEM Computational Labels and Weak Mechanics-Informed Regularization

Abstract: Systematic discrepancies between experimentally measured and numerically predicted strains can limit model validation and subsequent data-driven analysis. This study develops an experimental strain correction framework for a multi-point loaded cantilever beam using differential quadrature element method (DQEM) computational labels, ANSYS finite element validation, and weak mechanics-informed regularization. A total of 258 loading cases were analyzed at five strain-measurement locations. The DQEM labels showed good agreement with ANSYS, with a full-sample MAE of 1.6694 με. A point-wise correction model was then trained using experimental variables and a weak Euler–Bernoulli bending regularization term. On a load-case-level test set, the proposed model achieved an MAE of 1.8447 με and an RMSE of 5.8143 με. Compared with the uncorrected experimental data, the corrected results substantially reduced strain error while maintaining stable performance under noise perturbation. The results demonstrate that validated computational labels combined with weak mechanics-informed learning provide a practical approach for correcting systematic strain-measurement errors. Read More