Die chronologische Liste zeigt aktuelle Veröffentlichungen aus dem Forschungsbetrieb der Hochschule Weihenstephan-Triesdorf. Zuständig ist das Zentrum für Forschung und Wissenstransfer (ZFW).
8 Ergebnisse
Fabian Weckesser,
Prof. Dr. Frank Leßke,
Dipl. Ing. (FH) Marco Luthardt,
Prof. Dr. Kurt-Jürgen Hülsbergen
Data that are required for nutrient management are becoming increasingly available in digital format, leading to a high innovation potential for digital nitrogen (N) management applications. However, it is currently difficult for farmers to analyze, assess, and optimize N flows in their farms using the existing software. To improve digital N management, this study identified, evaluated, and systematized the requirements of stakeholders. Furthermore, digital farm N management tools with varying objectives in terms of system boundaries, data requirements, used methods and algorithms, performance, and practicality were appraised and categorized. According to the identified needs, the concept of a farm N management system (FNMS) software is presented which includes the following modules: (1) management of site and farm data, (2) determination of fertilizer requirements, (3) N balancing and cycles, (4) N turnover and losses, and (5) decision support. The aim of FNMS is to support farmers in their farming practices for increasing N efficiency and reducing environmentally harmful N surpluses. In this study, the conceptual requirements from the agricultural and computer science perspectives were determined as a basis for developing a consistent, scientifically sound, and user-friendly FNMS, especially applicable in European countries. This FNMS enables farmers and their advisors to make knowledge-based decisions based on comprehensive and integrated data.
Metabolomics approaches provide a vast array of analytical datasets, which require a comprehensive analytical, statistical, and biochemical workflow to reveal changes in metabolic profiles. The biological interpretation of mass spectrometric metabolomics results is still obstructed by the reliable identification of the metabolites as well as annotation and/or classification. In this work, the whole Lemna minor (common duckweed) was extracted using various solvents and analyzed utilizing polarity-extended liquid chromatography (reversed-phase liquid chromatography (RPLC)-hydrophilic interaction liquid chromatography (HILIC)) connected to two time-of-flight (TOF) mass spectrometer types, individually. This study (introduces and) discusses three relevant topics for the untargeted workflow: (1) A comparison study of metabolome samples was performed with an untargeted data handling workflow in two different labs with two different mass spectrometers using the same plant material type. (2) A statistical procedure was observed prioritizing significant detected features (dependent and independent of the mass spectrometer using the predictive methodology Orthogonal Partial Least Squares-Discriminant Analysis (OPLS-DA). (3) Relevant features were transferred to a prioritization tool (the FOR-IDENT platform (FI)) and were compared with the implemented compound database PLANT-IDENT (PI). This compound database is filled with relevant compounds of the Lemnaceae, Poaceae, Brassicaceae, and Nymphaceae families according to analytical criteria such as retention time (polarity and LogD (pH 7)) and accurate mass (empirical formula). Thus, an untargeted analysis was performed using the new tool as a prioritization and identification source for a hidden-target screening strategy. Consequently, forty-two compounds (amino acids, vitamins, flavonoids) could be recognized and subsequently validated in Lemna metabolic profile using reference standards. The class of flavonoids includes free aglycons and their glycosides. Further, according to our knowledge, the validated flavonoids robinetin and norwogonin were for the first time identified in the Lemna minor extracts.
Mehr
Dr. Ludwig Gredmaier,
Prof. Dr. Sabine Grüner-Lempart,
Julian Eckert,
Rainer Joachim,
Peter Funke
This is a knowledge contribution to the unsatisfactory biodegradation problem, when biotrickling filters are purifying mixed paint solvents. A biotrickling filter manufacturer reported low biodegradation rates during the purification of a hydrocarbon pollutant mix from an industrial paint spraying floor. From a gas chromatograph/mass spectrometer analysis both hydrophilic and hydrophobic solvents were found in the polluted air. It is known that biodegradation is retarded, if the pollutant does not transfer from gas to liquid into the biofilm and it was therefore suspected that hydrophobic pollutants do not sufficiently migrate into the water/biofilm. To test this hypothesis, pure, rather than mixed pollutants, were injected into the abiotic biotrickling filter. When hydrophobic paint solvent (xylene) was sprayed into the biotrickling filter, the solvent load at the outlet of the filter was almost as high as at the inlet. But when pure, hydrophilic paint solvent (PGME) was sprayed into the abiotic biotrickling filter, the solvent load measured at the outlet of the filter was zero, indicating complete dissolution into the circulation water. Carbon/solvent loads at the filter outlet and inlet were measured with a portable flame ionization detector instrument. The experiment confirms that the hydrophobic solvent does not migrate into the liquid phase. This poor mass transfer of hydrophobic solvents is likely to be the reason for the low biodegradation rate. The result is highly relevant to the paint spraying industry and manufacturers of exhaust gas treatment equipment alike, who spend millions in non-sustainable incineration of exhaust gases.
Human learning and intelligence work differently from the supervised pattern recognition approach adopted in most deep learning architectures. Humans seem to learn rich representations by exploration and imitation, build causal models of the world, and use both to flexibly solve new tasks. We suggest a simple but effective unsupervised model which develops such characteristics. The agent learns to represent the dynamical physical properties of its environment by intrinsically motivated exploration and performs inference on this representation to reach goals. For this, a set of self-organizing maps which represent state-action pairs is combined with a causal model for sequence prediction. The proposed system is evaluated in the cartpole environment. After an initial phase of playful exploration, the agent can execute kinematic simulations of the environment’s future and use those for action planning. We demonstrate its performance on a set of several related, but different one-shot imitation tasks, which the agent flexibly solves in an active inference style.
Mehr
B.Sc. Nikolas Trimpe,
Dr. Jörg Schäffer,
Prof. Dr. Sabine Grüner-Lempart
Bei zahlreichen industriellen Prozessen fallen lösemittelhaltige Abluftströme an, die zur Einhaltung der gesetzlich vorgeschriebenen Emissionsgrenzwerte gereinigt werden müssen. Konventionelle Abluftreinigungsverfahren wie die thermische Nachverbrennung erfordern jedoch hohe Temperaturen von 1200 °C und oft fossile Brennstoffe. Als deutlich nachhaltiger erweisen sich dagegen Biorieselbettreaktoren, mit denen Lösemittel bereits bei 20 °C biologisch abgebaut werden können. Die Schadstoffe werden dabei von Bakterien in einem Biofilm auf Lavasteinen in unbedenkliche Substanzen umgewandelt.
Mehr
Betreuung der Publikationsseiten
Gerhard Radlmayr
Referent für Wissenstransfer und Forschungskommunikation
Wir verwenden Cookies. Einige sind notwendig für die Funktion der Webseite, andere helfen uns, die Webseite zu verbessern. Um unseren eigenen Ansprüchen beim Datenschutz gerecht zu werden, erfassen wir lediglich anonymisierte Nutzerdaten mit „Matomo“. Um unser Internetangebot für Sie ansprechender zu gestalten, binden wir außerdem externe Inhalte unserer Social-Media-Kanäle ein.