https://scidar.kg.ac.rs/handle/123456789/8211 Fri, 15 May 2026 01:05:37 GMT 2026-05-15T01:05:37Z https://scidar.kg.ac.rs/handle/123456789/23131 Title: Multi-criteria calibration of a welding heat-transfer model: balancing geometric and temperature-based objectives Authors: Bjelić, Mišo; Radičević, Branko; Petrović, Aleksandra; Rasinac, Mladen; Mijajlovic, Miroslav Editors: Bakic, Vukman Abstract: Calibration of heat-source parameters and boundary heat losses is essential for reliable numerical welding simulations. This paper applies Grey relational analysis with entropy-based weighting to calibrate a three-dimensional heat-transfer model for gas metal arc welding of P355GH steel. Eight parameters are calibrated simultaneously: five describing the Goldak double-ellipsoidal heat source and three describing convective and radiative losses. Latin hypercube sampling with 80 combinations is used to explore the parameter space, and five objective functions are defined: three geometry-based and two temperature-based. Three calibration scenarios are compared: geometry-based, temperature-based, and combined, revealing a pronounced trade-off between geometric and thermal criteria. The combined scenario yields physically plausible parameter values and balanced prediction quality for both fusion-zone geometry and peak temperature indicators, and is therefore recommended for practical calibration of welding heat-transfer models. Description: TS_2026 - Bjelić M. et al. Thu, 01 Jan 2026 00:00:00 GMT https://scidar.kg.ac.rs/handle/123456789/23131 2026-01-01T00:00:00Z https://scidar.kg.ac.rs/handle/123456789/23126 Title: Development of a systematic instrumentation methodology for continuous torque measurement in high-capacity shafts Authors: Bižić, Milan; Petrović, Dragan Abstract: This paper presents the development of a systematic instrumentation methodology for continuous torque measurement in high-capacity shafts, based on the application of strain gauges. Within the proposed methodology, systematically derived solutions to the most important challenges in the development of instrumented shafts for torque measurement are presented, including the determination of optimal strain gauge locations, their arrangement and number, as well as the configuration of their connection into Wheatstone bridges. The main objective of the proposed methodology is to achieve high sensitivity and accuracy of torque measurement while simultaneously neutralizing parasitic effects in the output signal of the Wheatstone bridge. In addition, the model-based calibration procedure of the measurement system and an inverse torque determination algorithm based on the output signal of the Wheatstone bridge are presented. The developed methodology was tested on a finite element method (FEM) model of a high-capacity wind turbine shaft under various loading conditions, considering the combined effects of torque (to be measured), bending moments, as well as axial and transverse forces. The obtained results demonstrate that the developed methodology enables high numerical consistency in torque determination, with deviations between the prescribed torque in the FEM model and the torque obtained by inverse determination being less than 0.04%. Thus, it is confirmed that the proposed methodology can be successfully applied in the development of high-capacity instrumented shafts intended for continuous torque measurement. Thu, 01 Jan 2026 00:00:00 GMT https://scidar.kg.ac.rs/handle/123456789/23126 2026-01-01T00:00:00Z https://scidar.kg.ac.rs/handle/123456789/23124 Title: Dynamic evolution-driven domain adaptation with wavelet dual-path structure for cross-domain fault diagnosis Authors: Shen, Huikang; Sun, Yawei; Tao, Hongfeng; Stojanović, Vladimir Abstract: Traditional fault diagnosis methods often suffer from performance degradation under new working conditions due to distribution shifts between the source and target domains. To bridge this gap in cross-domain fault diagnosis (CDFD), the domain adaptation (DA) technique leverages transfer learning to align feature distributions, which facilitates knowledge transfer from labeled source domains to unlabeled target domains. Although existing studies on DA have demonstrated efficacy, they still face significant challenges due to abrupt domain shifts and insufficient feature discrimination. To overcome these problems, this study proposes a dynamic evolution mechanism to construct a sequence of hybrid domains that gradually evolves from the source to the target domain. This strategy establishes a smooth transition path to mitigate abrupt domain shifts. Additionally, a dual-path feature extraction structure empowered by wavelet packet transform (WPT) is introduced. This structure decomposes input signals into high-frequency and low-frequency components to enhance discriminative feature representation. The experimental results on rolling bearing and gearbox datasets demonstrate the effectiveness and generalization performance of the proposed method. Thu, 01 Jan 2026 00:00:00 GMT https://scidar.kg.ac.rs/handle/123456789/23124 2026-01-01T00:00:00Z https://scidar.kg.ac.rs/handle/123456789/23113 Title: High Dimensional Spatial Information and Multi-scale Fusion Network for Efficient and Real-Time Small Object Detection in Remote Sensing Images Authors: Li, Haochen; Tao, Hongfeng; Qiu, Jier; Stojanović, Vladimir Abstract: The detection of small objects in remote sensing imagery remains a formidable challenge due to their minimal pixel occupancy, blurred structural boundaries, and susceptibility to environmental interference. To solve these problems, this paper proposes a novel network architecture named multidimensional information feature fusion-you only look once (MIFF-YOLO), which integrates several specialized modules. To address the challenge of small objects being obscured by complex environmental factors, we propose a multidimensional information fusion (MIF) module for the neck network, which leverages a 3D convolution and a full-domain transformer (FDT) to create cross scale dependencies and integrate global contextual information with local details. For the purpose of preserving the spatial and edge information of small objects, an efficient front end module (EFEM) is embedded into the C3k2 architecture. The EFEM module employs a parallel, learnable dual-path architecture that collaboratively integrates a Sobel convolution stream for explicit edge detection and a spatial information stream max-pooling for detail preservation, enabling simultaneous extraction of structural boundaries and contextual textures. These complementary features undergo an adaptive fusion via omni-dimensional dynamic convolution (ODConv), thereby enriching the capabilities of the feature representation. In order to address the loss of critical details in small object features during enlargement, dynamic upconvolution block (DUB) is introduced to replace standard upsampling module. Adaptive feature sampling is achieved through content-aware dynamic offsets, mitigating detail loss during resolution recovery. Compared with the original baseline algorithm, the improved network achieved a 3.7% improvement on mAP@50 and a 3.9% improvement on mAP@50:95, with the FPS reaching 120 on the DOTA dataset. This shows that the improved algorithm effectively enhances small object detection performance in remote sensing images while maintaining excellent real-time detection efficiency. Thu, 01 Jan 2026 00:00:00 GMT https://scidar.kg.ac.rs/handle/123456789/23113 2026-01-01T00:00:00Z