@techreport{OPUS4-7289, type = {Working Paper}, title = {Bazar Abbas : a recovery lab for Port Said ; built heritage, collective memory, and urban rehabilitation}, editor = {Zarrin Ghalam, Sepideh}, isbn = {2941-7406}, doi = {10.26127/BTUOpen-7289}, url = {http://nbn-resolving.de/urn:nbn:de:kobv:co1-opus4-72895}, year = {2026}, abstract = {This publication documents the processes and findings of the 'Bazar Abbas: Recovery Lab of Port Said', a pilot initiative addressing the deterioration of Port Said's historic urban landscape. By presenting these results, the publication aims to raise awareness of the city's urban heritage and stimulate a broader discussion on sustainable recovery strategies. Founded as a cosmopolitan hub after the construction of the Suez Canal, Port Said's dual urban structure represents a distinctive architectural and urban heritage. Founded as a cosmopolitan hub after the construction of the Suez Canal, Port Said's dual urban structure represents a distinctive, yet vulnerable, architectural and urban heritage. The city's urban fabric was profoundly shaped by successive wars and forced displacement following the post-colonial transition. Despite the profound impact of these events, however, Port Said has lacked a comprehensive recovery framework to address the resulting damage. Instead, the historic core has faced institutional neglect, with its heritage caught between shifting priorities and chronic underfunding. Today, the city is grappling with historical trauma and physical decay, as well as a development strategy that disconnects heritage from contemporary growth. In response, the Recovery Lab acts as a platform for exploration, bridging the gap between academic research and professional practice. Using the Bazar Abbas building as a living laboratory, the project has evolved from theoretical study to the implementation of site-specific pilot interventions. The Lab has tested a multidisciplinary toolbox ranging from 3D laser scanning to participatory engagement to demonstrate how tangible recovery can be achieved despite limited resources and archival gaps. Ultimately, this volume argues that cities with complex, layered histories require an incremental, tactical approach to rehabilitation. This methodology directly addresses intertwined challenges, including restricted funding, fragmented decision-making, and administrative hurdles. By treating heritage as a vital social asset and adopting a learning-by-doing approach, the Lab overcomes institutional paralysis to deliver results. This framework offers a resilient regeneration alternative that is grounded in the practical realities of Port Said and the collective agency of its residents.}, subject = {Port Said; Bazar Abbas; Stadterneuerung; Gemeinschaftsorientierte Sanierung; Urban recovery; Urban narratives; Community-centred rehabilitation; Port Said; Stadtsanierung; Stadtkern; St{\"a}dtebau; Geschichte; Urbane Narrative}, language = {en} } @phdthesis{Alfeus2026, author = {Alfeus, Martha}, title = {Rangeland improvement and biodiversity conservation through bush encroachment control in Southern Africa, Namibia}, doi = {10.26127/BTUOpen-7306}, url = {http://nbn-resolving.de/urn:nbn:de:kobv:co1-opus4-73064}, school = {BTU Cottbus - Senftenberg}, year = {2026}, abstract = {Biodiversity underpins ecosystem stability, resilience, and the delivery of ecological services essential to human well-being. However, global biodiversity is declining at unprecedented rates due to land-use change, habitat degradation, and climate variability. In arid and semi-arid savannahs, where ecosystems are constrained by resource scarcity these pressures intensify through bush encroachment, the densification of woody vegetation displacing grasslands. This shift undermines rangeland productivity, alters vegetation structure, and reshapes biodiversity patterns. While vegetation and livestock responses are well-studied, implications for invertebrate biodiversity, particularly arthropods as functionally critical taxa, remain underexplored. This thesis bridges these gaps via three interconnected objectives linking data integrity, ecological patterns, and management interventions in Namibian savannahs. The first assesses arthropod biodiversity data quality across global platforms (iNaturalist, GBIF) and the Namibian Biodiversity Database. Recognition (species-level identifications) and completeness (species coverage) vary markedly by arthropod order, influenced by detectability, morphology, and familiarity. Global datasets provide baselines, but national databases are vital for accuracy, advocating integrated frameworks to strengthen monitoring, conservation, and citizen science. The second study investigates effects of dominant encroaching shrubs, Dichrostachys cinerea, Senegalia mellifera, and Terminalia sericea on arthropod communities across canopy, aerial, and ground strata in wet/dry seasons. Sampling via pitfall, pan, and beating traps revealed shrub-specific compositional shifts, driven by canopy taxa. S. mellifera favoured omnivore guilds and D. cinerea favoured phytophagous. Abundance increased with ground cover but fell with woody density, highlighting habitat heterogeneity's role in sustaining diversity amid encroachment. The third objective evaluates bush thinning management practice: manual/chemical treatments at 50\% (moderate) and 100\% (complete) intensities. Moderate manual thinning boosted arthropod abundance, richness, and functional diversity versus complete/chemical options. Pollinators, herbivores, decomposers, and predators thrived under partial shrub thinning, preserving heterogeneity; while chemical clearing suppressed sensitive groups like Auchenorrhyncha, Opiliones, and Coccinellidae. Soil properties, including pH and nutrient concentrations, further modified arthropod responses across shrub species. In summary, the thesis presents a framework for arthropod conservation based on three pillars: robust data systems, vegetation-arthropod ecology, and balanced rangeland management that restores grasslands (e.g. 50\% moderate manual shrub thinning) without causing arthropod biodiversity loss. Integrating informatics, community ecology, and restoration advances dryland resilience in African savannahs and rangeland ecosystems.}, subject = {Arthropod biodiversity; Citizen science; Rangeland restoration; Shrub thinning; Namibia biodiversity; Arthropoden; Biodiversit{\"a}t; Namibia; Weidefl{\"a}chenrestaurierung; Strauchverj{\"u}ngung}, language = {en} } @phdthesis{ReyesPacheco2026, author = {Reyes Pacheco, Juan-Francisco}, title = {Explaining large language models for passage-level political statement extraction using linguistic rule-based models}, doi = {10.26127/BTUOpen-7293}, url = {http://nbn-resolving.de/urn:nbn:de:kobv:co1-opus4-72930}, school = {BTU Cottbus - Senftenberg}, year = {2026}, abstract = {The complexity of processing the language in political discourse presents a challenge due to its ambiguity and nuance. However, the growing volume of digital political content and the proven efficiency of Large Language Models (LLMs) in processing and "understanding" such statements presents an opportunity to explore new approaches. Therefore, the automatic extraction of coherent political statements at the passage level is an opportune task in the current era of increasing digitization of political discourse. This research addresses the problem of accurately extracting "relevant" political statements. Moving beyond traditional rule-based information extraction (IE) techniques, we propose a Natural Language Processing (NLP) pipeline composed of state-of-the-art LLMs. Given the opaque nature of Transformer models and the need for AI transparency — moreover, within the intricate nature of political language — this thesis develops an approach based on Explainable Artificial Intelligence (XAI) tools to explain LLM classification behavior from a linguistic perspective across the pipeline's NLP tasks. The empirical approach is structured around three specific NLP challenges, addressed in dedicated papers: (1) Text Genre Classification, employing a BERT model to distinguish between speeches and interviews based on the number of speakers, thereby addressing text genre identification and filtering; (2) Stance Classification, utilizing a few-shot learning model (SetFit) for binary classification of stance expressions, enhancing the detection of relevant political statements; and (3) Topic Continuity Analysis, using Sentence Pair Modeling (SPM) and BERT to analyze topic continuity between sentence pairs, ensuring the coherence of extracted passages. This research enhances explainability across these challenges by combining linguistic rule-based models (LRBMs) with contemporary explainability tools like SHAP and Transformers Interpret. By leveraging corpus linguistics, tailored lexicons, and lexicogrammatical rules, our research adds transparency to LLMs' decision-making processes, bridging traditional linguistic theories with modern neural models. Likewise, although not the primary focus, our research emphasized model performance, consequently increasing the accuracy of their explanatory capacity. For each challenge, ad hoc datasets focused on American political discourse in English were created. LRBMs played a crucial role in feature engineering and dataset annotation, further improving the explainability and performance of XAI tools. Fusing linguistic insight with neural outputs allowed for granular token-level analysis of morphosyntactic, semantic, and pragmatic features, offering a novel contribution by aligning outputs with established linguistic theories. This thesis advances the understanding of LLM explainability by approaching the extraction of relevant political statements from a linguistically grounded framework.}, subject = {Large language models; Information extraction; NLP; Politics; Discourse analysis; XAI; Politik; Diskursanalyse; LLM; Explainable artificial intelligence; Politik; Diskursanalyse; Großes Sprachmodell; Automatische Sprachanalyse; Erkl{\"a}rbare K{\"u}nstliche Intelligenz}, language = {en} } @phdthesis{Hady2025, author = {Hady, Mohamed Elfatih Ahmed Omer}, title = {Stochastic optimal control of occupational pension funds}, doi = {10.26127/BTUOpen-7286}, url = {http://nbn-resolving.de/urn:nbn:de:kobv:co1-opus4-72867}, school = {BTU Cottbus - Senftenberg}, year = {2025}, abstract = {In this dissertation, we consider occupational pension funds that are set up by a company for its employees. A special feature of such pension funds is that the collectives of insured individuals are typically smaller, and benefit payments depend on the seniority and salary of the insured. Therefore, fluctuations over time in the composition of the collective of insured persons in terms of age, length of service, and salary cannot be ignored when calculating the actuarial liabilities and the company's financial contributions to the fund. We describe the stochastic dynamics of the composition of the collective of insured individuals by a discrete-time Markov chain model. This allows us to derive the dynamics of the expected aggregated actuarial liabilities for all insured individuals of the collective as a functional of the Markov chain. They are then approximated by mean-reverting processes, which facilitate the formulation of stochastic optimal control problems that arise in the cost-optimal management of occupational pension funds. Here, the objective is to invest the fund's capital in a financial market such that the expected net present value of all future contributions to the fund is minimized. The resulting optimal funding problem is studied under additional constraints that capture the obligations imposed by a regulator. They are intended to ensure that the fund capital is always sufficient to finance current benefit payments and future actuarial obligations. This problem is analyzed firstly under additional dynamic risk constraints and secondly under tracking constraints, which are included in the performance criterion with the help of penalties. While the risk constraints aim to keep the fund assets above the actuarial liabilities, the tracking constraints aim to keep these assets as close as possible to the liabilities. The resulting discrete-time stochastic optimal control problems are treated as Markov decision processes (MDPs). The corresponding Bellman equation for the value function is solved numerically using a backward recursion algorithm. We present the results of extensive numerical experiments showing the properties of the value function and the optimal decision rules, as well as their dependence on the various risk constraints.}, subject = {Occupational pension fund; Stochastic optimal control; Markov chain model; Actuarial liabilities; Dynamic risk constraints; Betriebliche Altersvorsorge; Stochastische optimale Steuerung; Markovketten-Modell; Versicherungsmathematischen Verpflichtungen; Dynamische Risikobeschr{\"a}nkung; Betriebliche Altersversorgung; Versicherungsmathematik; Markov-Kette}, language = {en} } @article{PillingBaehrWunderlich2026, author = {Pilling, Eric and B{\"a}hr, Martin and Wunderlich, Ralf}, title = {Reinforcement learning methods for the stochastic optimal control of an industrial power-to-heat system}, series = {Energies}, volume = {19}, journal = {Energies}, number = {4}, publisher = {MDPI}, address = {Basel}, issn = {1996-1073}, doi = {10.3390/en19041046}, year = {2026}, abstract = {The optimal control of sustainable energy supply systems, including renewable energies and energy storage, takes a central role in the decarbonization of industrial systems. However, the use of fluctuating renewable energies leads to fluctuations in energy generation and requires a suitable control strategy for the complex systems in order to ensure energy supply. In this paper, we consider an electrified power-to-heat system which is designed to supply heat in the form of superheated steam for industrial processes. The system consists of a high-temperature heat pump for heat supply, a wind turbine for power generation, a sensible thermal energy storage for storing excess heat, and a steam generator for providing steam. If the system's energy demand cannot be covered by electricity from the wind turbine, additional electricity must be purchased from the power grid. For this system, we investigate the cost-optimal operation, aiming to minimize the electricity cost from the grid by a suitable system control depending on the available wind power and the amount of stored thermal energy. This is a decision-making problem under uncertainty regarding the future prices for electricity from the grid and the future generation of wind power. The resulting stochastic optimal control problem is treated as finite-horizon Markov decision process for a multi-dimensional controlled state process. We first consider the classical backward recursion technique for solving the associated dynamic programming equation for the value function and compute the optimal decision rule. Since that approach suffers from the curse of dimensionality, we also apply reinforcement learning techniques, namely Q-learning, that are able to provide a good approximate solution to the optimization problem within reasonable time.}, subject = {Stochastic optimal control; Markov decision process; Dynamic programming; Q-learning; Power-to-heat system}, language = {en} } @article{DjoudiAmosseAvironetal.2026, author = {Djoudi, El Aziz and Amoss{\´e}, Ewen and Aviron, St{\´e}phanie and Birkhofer, Klaus and Libereau, Benjamin and Plantegenest, Manuel and P{\´e}tillon, Julien}, title = {Multi‑trophic arthropod communities modulated by local farming system and landscape heterogeneity}, series = {Landscape Ecology}, volume = {41}, journal = {Landscape Ecology}, publisher = {Springer Nature}, address = {Berlin}, issn = {1572-9761}, doi = {10.1007/s10980-026-02298-y}, year = {2026}, abstract = {Organic farming is increasingly used worldwide in recent decades, but the role of this farming system in shaping the trophic structure of arthropod communities remains poorly understood, especially at different spatial scales. Therefore, the contribution of landscape heterogeneity in shaping arthropod's trophic guilds remain understudied. In this study, we assessed how the evenness, the abundance and the taxonomic richness of arthropod trophic groups were shaped by the local farming system, landscape heterogeneity (including the percentage of semi-natural habitats and organic fields), and their interaction. Arthropod's trophic guilds (ground and vegetation-dwelling) were sampled in 20 spatial independent pairs of conventional (CF) and organic (OF) fields located in Brittany (Western France) by using pitfall traps and sweep nets replicated in time and space. A total of 95,822 arthropods belonging to 197 taxonomic groups were sampled. Farming system has a strong overall positive effect on the community structure of both ground- and vegetation-dwelling arthropods. Landscape heterogeneity, alone and in interaction with farming systems, affected positively the diversity and abundance of most trophic groups for both ground- and vegetation-dwelling arthropods. Organic farming therefore affected the trophic composition of arthropod communities not only locally, but also at the landscape scale. Our study highlights the strong positive effect of farming system and the landscape heterogeneity on arthropod communities and more importantly the interconnection between different spatial scales in this process. Taking these aspects into account is therefore crucial in understanding arthropod community dynamics in agroecosystems.}, subject = {Agroecology; Agricultural geography; Biodiversity; Landscape ecology; Organic farming}, language = {en} } @phdthesis{MenaresBarraza2025, author = {Menares Barraza, Esteban Javier}, title = {Beyond species richness : integrating ecological interactions and multiple goals in grassland conservation}, doi = {10.26127/BTUOpen-7233}, url = {http://nbn-resolving.de/urn:nbn:de:kobv:co1-opus4-72336}, school = {BTU Cottbus - Senftenberg}, year = {2025}, abstract = {This thesis explores how incorporating trophic interactions and a wider range of conservation goals can enhance the design and implementation of biodiversity conservation measures in permanent grasslands. Based on data from two long-term biodiversity monitoring regions in Germany [Schw{\"a}bische Alb (ALB) and Schorfheide-Chorin (SCH)], the thesis is structured around three main chapters. Together, these chapters examine (1) the reliability of co-occurrence methods for inferring plant-butterfly interactions, (2) the cost-effectiveness of different land-use management practices for achieving multiple conservation goals, and (3) the detection of ecological thresholds for key drivers of change in these goals at species, community, and ecosystem levels. I employed a combination of literature review, field data, statistical inference models, optimisation algorithms, and multivariate analyses to provide a thorough evaluation of grassland biodiversity and move beyond traditional species richness metrics by incorporating interaction networks and other metrics, thus capturing ecological complexity relevant to conservation planning. Chapter I shows that, despite the Relative Interaction Intensity (RII) with pairwise null models approach performing better than the other methods tested, co-occurrence methods fail to reliably predict known plant-butterfly interactions. This highlights the limited usefulness of such methods for mobile species and the importance of directly observing trophic interactions. Chapter II shows that optimising land-use management for multiple conservation goals under cost-effectiveness can achieve substantial ecological benefits (up to 80\% of the maximum) using as little as 30\% of the budget. Although trade-offs were observed, certain goals, including trophic interactions, butterfly richness, and network nestedness, showed more synergies. Optimising for sets of goals consistently outperformed single-goal strategies, though regional variation in optimal management highlights the need for region-specific planning. Chapter III presents ecological thresholds for key biodiversity goals along management and landscape gradients. In ALB, most goals showed sharp declines above one to two cuts per year, and low grazing levels were associated with higher multidiversity and more complex interaction networks. In SCH, network connectance and the number of interactions increased even at low levels of forest cover (2-10\%), highlighting the importance of maintaining forest patches near grasslands for supporting biodiversity. Additionally, even low fertiliser inputs were associated with species-poor butterfly communities composed mainly of common species with wider distribution ranges. These results provide key insights for improving region-specific management strategies in conservation programmes and show the importance of using multiple conservation goals, especially those capturing species interactions, when designing effective conservation strategies.}, subject = {Co-occurrence analysis ; Trophic interactions; Network metrics; Cost-effective optimisation; Trade-off analysis; Kookkurrenzanalyse <{\"O}kologie>; Trophische Interaktionen; Netzwerkkennzahlen; Kosteneffektive Optimierung; Trade-off-Analyse; Biodiversit{\"a}t; Naturschutzmonitoring; Effizienz; Artengruppe; Nahrungskette; Schmetterlinge}, language = {en} } @article{MbuyaPawarJafarietal.2025, author = {Mbuya, Christel Olivier Lenge and Pawar, Kunal and Jafari, Mitra and Shafiee, Parisa and Okoye Chine, Chike George and Tarifa, Pilar and Dorneanu, Bogdan and Arellano-Garcia, Harvey}, title = {Tuning catalyst performance in methane dry reforming via microwave irradiation of nickel-silicon carbide systems}, series = {Journal of CO2 Utilization}, volume = {102}, journal = {Journal of CO2 Utilization}, publisher = {Elsevier}, address = {Amsterdam}, issn = {2212-9820}, doi = {10.1016/j.jcou.2025.103270}, url = {http://nbn-resolving.de/urn:nbn:de:kobv:co1-opus4-72832}, year = {2025}, abstract = {The dry reforming of methane (DRM) is a promising route for converting greenhouse gases such as methane (CH4) and carbon dioxide (CO2) into valuable syngas, hydrogen (H2) and carbon monoxide (CO). However, traditional nickel (Ni)-based catalysts suffer from rapid deactivation due to carbon deposition and sintering, especially when supported on low thermal conductivity materials. In this work, a novel post-synthesis microwave irradiation (MIR) treatment is introduced to systematically optimize the performance of Ni - β - SiC and Ni - Ti - Cβ - SiC catalysts for DRM. Unlike previous studies that have used MIR during reaction or with different supports, this approach tunes the metal - support interactions and textural properties of Ni - β - SiC and Ni - Ti - Cβ - SiC catalysts by varying the MIR exposure time after catalyst synthesis. MIR post-treatment (10-25 s) increased the CH4 conversion to 65 \% and the CO2 conversions to 62 \% for Ni-β-SiC catalysts and improved the H₂/CO ratio to 0.80, with stable performance over 20 h. For Ni-Ti-Cβ-SiC, MIR (10-20 s) maintained CH4 conversion up to 60 \% and CO2 conversion to 58 \% over 20 h, while the untreated catalyst, though initially higher, deactivated rapidly. Excessive MIR (30 s) reduced performance for both catalyst types, underscoring the need for optimal exposure time. These findings demonstrate post-synthesis MIR provides a tuneable approach for enhancing both the activity and durability of Ni/SiC - based DRM catalysts through controlled modification of metal - support interactions. This work offers new insights for the design of robust catalysts aimed at greenhouse gas utilization and sustainable syngas production, with activity and stability enhancements linked to controlled changes in metal - support interactions.}, subject = {Carbon dioxide; Dry reforming; Methane; Microwave irradiation; Ni Silicon carbide catalysts}, language = {en} } @article{SharmaRadomskyMathiazhaganetal.2025, author = {Sharma, Dikshant and Radomsky, Lukas and Mathiazhagan, Akilan and Asli, Majid and H{\"o}schler, Klaus and Mallwitz, Regine}, title = {Thermal analysis of metal foam integrated heatsink for electrified aircraft applications}, series = {International Journal of Thermofluids}, volume = {30}, journal = {International Journal of Thermofluids}, publisher = {Elsevier}, address = {Amsterdam}, issn = {2666-2027}, doi = {10.1016/j.ijft.2025.101465}, year = {2025}, abstract = {Metal foams facilitate large heat dissipation in high-power dense systems such as power electronics for electrified propulsion application. This work addresses the cooling of a power semiconductor device with aluminium and copper metal foam integrated hybrid heatsink and its comparative analysis to a conventional finned heatsink using 0D and 3D modelling approach. Two-equation foam model in Fluent is utilized and the numerical approach is validated against experimental dataset. Inlet air velocity is varied such that the Darcy to turbulent regimes of the open-cellular foams are covered. Foam porosities from ∼ 0.85-0.95 with 10-20 PPI are investigated and the thermal performance of the heatsink is found to be independent of the foam material. High porosity (> 0.9), 20 PPI foams are found to aid forced-convection by improving the thermal resistance by more than 10\% against the 10 PPI counterparts. The hybrid heatsink outperforms the foam-based and conventional heatsink by 40\% and 15\% respectively when considering the reduction in junction temperatures, while the hydraulic resistance increases 10 times when compared to the conventional one. The 0D thermal resistance model is robust in predicting the junction temperatures for metal foam heatsinks with only a 5\%-6\% discrepancy for both the 50 W and 100 W heat load scenarios. The key and novel contribution of this study is the integration of detailed 3D simulations of a power electronics cooling environment with the development of a corresponding 0D thermal model. This approach not only eases the physical representation of the system but also enables the model to be extended to diverse heat load conditions.}, subject = {Electric propulsion; Electronics thermal management; Metal foam heatsink; Porous media; Forced convection}, language = {en} } @article{YangHenkeMoehlenkamp2025, author = {Yang, Yinghui and Henke, Markus and M{\"o}hlenkamp, Georg}, title = {Analytical investigation of short-circuit protection and post-fault operation in dual three-phase PMSMs for hybrid electric counter-rotating compressor applications}, series = {IEEE Access}, volume = {14}, journal = {IEEE Access}, publisher = {IEEE}, address = {New York}, issn = {2169-3536}, doi = {10.1109/ACCESS.2026.3658195}, pages = {17303 -- 17323}, year = {2025}, abstract = {The hybrid electric counter-rotating compressor (CRC) has the potential to improve power-specific fuel consumption in aircraft engines. However, due to its cascaded structure, a reduction or loss of driving power in a single compressor stage can significantly degrade the overall compressor performance. A short-circuit fault is a typical electrical failure in an electric motor that can lead to a complete loss of driving power. One innovative post-fault operation strategy is to utilize the braking capability of the failed motor at low speed to lock its corresponding compressor stage. In this way, the failed stage can function as a stator stage, still providing degraded flow-guiding capability and preventing reverse rotation. This article investigates the short-circuit protection strategy and post-fault operation of dual three-phase permanent magnet synchronous motors (DTP-PMSMs) for hybrid electric CRC applications. Analytical models are developed to characterize the post-fault behavior of the DTP-PMSM. Key design factors are defined, and their influence on post-fault performance is analyzed. Critical post-fault operating points are identified to evaluate and compare different protection strategies. Based on these results, an optimized shortcircuit protection strategy is proposed for DTP-PMSMs in CRC applications. Finally, the analytical findings and the proposed strategy are validated through simulation.}, subject = {Windings; Circuit faults; Rotors; Torque; Protection}, language = {en} }