Functional beverages enriched with herbal extracts are gaining popularity due to their potential health benefits. Tilia cordata flowers are known for their antioxidant and antimicrobial properties, making them a promising additive in food and beverage formulations. Our study aimed to develop ready-to-drink iced teas enriched with T. cordata flower extracts and to evaluate their antioxidant, antimicrobial, and sensory characteristics as functional food products. Fresh T. cordata flowers were analyzed for metal contents. Phenolic acid profiles in ethanolic and aqueous extracts were determined using HPLC-MS. Antioxidant activity was evaluated using DPPH radical scavenging, conjugated diene, and iron ion chelation assays. Antimicrobial effects were tested against Staphylococcus aureus, Bacillus cereus, and Listeria monocytogenes. Sensory analysis was conducted using AI-based facial expression recognition to assess consumer responses. Metal analysis revealed low concentrations of Mn, Zn, Cu, and Fe, with no detectable Pb, Cd, or Ni. Ethanolic extracts showed significantly higher levels of phenolic acids than aqueous extracts. Iced teas containing both types of extracts demonstrated strong antioxidant activity, with ethanolic formulations having the highest levels of phenols and flavonoids. Antimicrobial tests confirmed activity in both teas, with ethanolic extracts showing stronger effects. Sensory analysis indicated positive emotional responses and consumer acceptance for both formulations. Iced teas enriched with T. cordata extracts exhibited significant antioxidant and antimicrobial properties, confirming their potential as functional beverages. The use of AI-driven sensory evaluation proved effective in capturing consumer preferences, supporting its application in product development. These findings suggest commercial viability for industrial production.
The urban-suburban railway system of Belgrade currently accounts for a relatively small share of passenger transport in the metropolitan area and is undergoing substantial modernization from both technological and technical perspectives. This comprehensive effort includes not only infrastructure enhancements but also the renewal of the electric multiple unit (EMU) fleet. Defining procurement requirements for new EMUs is a complex decision-making task involving numerous interdependent and potentially conflicting attributes. To support this process, this paper applies two integrated methods: the Decision-Making Trial and Evaluation Laboratory combined with the Analytic Network Process (DEMATEL-ANP), and its advanced variant, picture fuzzy DEMATEL-ANP (PF-DEMATEL-ANP). The two approaches are used to identify and compare the relative importance and causal interdependencies of attributes relevant to EMU fleet renewal. Both approaches identify reliability- and availability-related attributes as the most influential factors in prioritizing EMU procurement requirements. The results of the Monte Carlo sensitivity analysis indicate a high level of stability in the resulting weights and rankings. The goal of this paper is to provide a clear and practical decision-support framework that helps railway managers and policy makers prioritize the most critical attributes relevant to the EMU procurement process, thereby supporting the subsequent evaluation of specific alternatives.
Rural settlements in Serbia are increasingly exposed to the combined pressures of demographic decline and climate-related hazards. This study integrates Geographic Information Systems (GIS), Remote Sensing (RS), census data, gridded population datasets, and climate projections to assess long-term rural vulnerability through 2100. Historical census information spanning 1948-2022 provides the demographic context, while the 1991 and 2022 rural populations are used for the baseline demographic projection. Landsat 8/9, Sentinel-2, hazard-event databases, and CMIP6-MIROC6 climate projections under SSP5-8.5, downscaled using CHELSA, are integrated in the spatial analysis. A dimensionless Climate-Hazard Exposure Index (CHEI), based on normalized drought, flood, extreme-rainfall, and temperature scores, and a Settlement Viability Index (SVI) are used to identify areas where demographic decline coincides with elevated climate-hazard exposure. The baseline demographic model estimates a reduction of the rural population from approximately 2.1 million inhabitants in 2022 to 1.44 million by 2050 and about 730,000 by 2100. Climate exposure is treated as an independent spatial layer rather than as a calibrated causal coefficient of population decline. Severe demographic viability risk is operationalized using a projected settlement population threshold of fewer than 100 inhabitants. The spatial classification achieved an overall accuracy of 80.0%, with precision of 83.0%, recall of 80.0%, and an F1-score of 81.5%. Demographic hindcasting is interpreted as an internal agreement check rather than independent predictive validation. Southern and eastern Serbia show the highest combined vulnerability, supporting the need for targeted adaptation, climate-resilient infrastructure, and sustainable rural-development policies.
This paper addresses the Longest Filled Common Subsequence (LFCS) problem, a challenging NP-hard problem with applications in bioinformatics, including gene mutation prediction and genomic data reconstruction. Existing approaches, including exact, metaheuristic, and approximation algorithms, have primarily been evaluated on small instances, which provide limited insight into their scalability. In this work, we introduce a new benchmark dataset with significantly larger instances and demonstrate that existing datasets lack the discriminative power needed to meaningfully assess algorithm performance at scale. To solve large instances efficiently, we utilize an adaptive Construct, Merge, Solve, Adapt (CMSA) framework that iteratively generates promising subproblems via component-based construction and refines them using feedback from prior iterations. Subproblems are solved using an external black-box solver. Extensive experiments on both standard and a newly introduced set of large-sized instances prove that the designed adaptive CMSA achieves state-of-the-art performance, outperforming four known approaches. Notably, among 1510 problem instances with known optimal solutions, our approach matches the proven optimum on 1486 instances—corresponding to 98.4% of these instances. These results demonstrate substantially improved scalability on large problem instances. As an engineering contribution, we propose a novel application of LFCS to song identification from degraded audio excerpts, using real-world energy-profile instances derived from popular music. Finally, we conduct an empirical explainability analysis to identify critical feature combinations influencing algorithm performance, revealing the key problem features associated with the success or failure of the approaches across different instance types.
This study investigates a sustainable route for biodiesel production from waste plum kernel oil (PKO) using a heterogeneous, waste-derived catalyst (35% CaO supported on coal fly ash-derived zeolite), in the presence of a green deep eutectic solvent (DES), triethanolamine:menthol (TEA:M, 1:2 mol/mol), as a cosolvent. The catalyst exhibited a high specific surface area, developed porosity, and abundant strong basic sites, ensuring effective transesterification activity. PKO, characterized by low acid value (1.95 +/- 0.03 mg KOH/g) and high oleic acid content (73.8 +/- 2.5%), proved to be a suitable feedstock for biodiesel production. The incorporation of TEA:M DES into the reaction system significantly enhanced fatty acid methyl ester (FAME) formation by improving mass transfer during the initial reaction period. Transesterification was conducted at 60 degrees C under atmospheric pressure by varying the catalyst amount (2-10% based on oil mass), DES amount (1-9% based on oil mass), and methanolto-oil molar ratio (6:1-12:1 mol/mol). A maximum FAME content of 96.7% was achieved under optimized reaction conditions (8% catalyst, 5% DES, 6:1 methanol-to-oil ratio). Transesterification followed a pseudo-firstorder kinetics with rate constants of 0.032-0.152 min(-1) and an activation energy of 42.67 kJ/mol. The catalyst maintained high activity over five reuse cycles, with only a slight decrease in FAME content from 96.7 +/- 1.0% to 94.5 +/- 0.5%. Cost analysis confirmed process feasibility ((sic)1.24/kg), demonstrating the potential of integrating waste-derived catalysts with green cosolvents for sustainable biodiesel production and providing a promising pathway toward a circular economy.
