Open Access

Das Force Concept Inventory ist ein international etabliertes Diagnoseinstrument für das Konzeptverständnis der Newtonschen Mechanik. Wir haben über einen Zeitraum von zehn Jahren von knapp 5000 Studierenden der Ingenieurwissenschaften an der TH Rosenheim zu Studienbeginn Testdaten erhoben. Die Auswertung mittels Item Response Curves erlaubt für jede Frage die Analyse sowohl der richtigen als auch aller falschen Antwortoptionen. Insbesondere lassen sich auf diese Weise Distraktoren identifizieren, die gängige Fehlvorstellungen widerspiegeln. Es zeigt sich, dass die am häufigsten gewählten Distraktoren auf der Annahme beruhen, jeder Bewegung müsse eine Kraft in Bewegungsrichtung zu Grunde liegen. Daten von Universitäten der USA zeigen ein vergleichbares Ergebnis. Die vorgestellte Analyse dient als Grundlage zur Verbesserung von Lehrsequenzen.

The applications for occupancy detection range from controlling building automation and systems, determining heat transfer coefficients and even assessing the risk of infection in rooms. Studies in the literature use various statistical models, physical models and machine learning techniques to detect occupancy. All these methods require data for training the occupancy detection models. However, data generation is time-consuming and expensive. This study demonstrates the feasibility of using simulated learning data. Using three different data sources, we tested the suitability of different methods for generating learning data. We conducted two experiments in two office spaces with a real user and an artificial user, and we generated a third data set using a building simulation model. In addition, this study compares two different machine learning approaches (Random Forest and LASSO) using environmental parameters. Both machine learning approaches could develop models with a sensitivity of at least 83 % and a specificity of at least 97 % with both training data sets. This work shows that it is possible to determine the presence in rooms using simulated data. The results compared to measured data were just slightly less accurate, and the added value due to the lower effort was considerable

To detect degradation of photovoltaic systems at an early stage, precise performance determinations are essential. Current-voltage measurements on single modules in indoor labs under well-defined conditions are state-of-the-art. However, this method causes a logistical and economic effort. On-site measurements reduce the effort but are prone to error because of poorly determinable test conditions: The detected temperature on the backside of modules as well as the irradiance detected by pyranometer in tilted module plane usually differs from the actual operating conditions. In consequence, the accuracy of measured current-voltage characteristics is poor. A precise on-site power determination is achieved by a self-reference algorithm. Improvements by self-referencing are not achieved by correcting the raw electrical data, but by replacing the measured test conditions (temperature and irradiance) with effective data. These effective values correspond better to the actual module irradiance and temperature values. Furthermore, precise power determination of photovoltaic arrays requires to identify and exclude emporarily deformed current-voltage curves from consideration, as it can arise from non-uniform irradiation within array like shading. In this paper, the application of a digital curve filter is reported and performance determinations by a self-reference algorithm are demonstrated on a photovoltaic array. We review and discuss the measurement concept in terms of its ease of use. …