This paper considers an adaptive orthogonal frequency-division multiplexing (OFDM) transmission scheme for narrowband (NB) power line communication (PLC). The NB PLC channel is found to show strong cyclostationary characteristics due to the periodic mains voltage. However, conventional systems do not use these time-varying channel features. An improvement of the performance is achieved by exploiting the cyclic properties of the channel. Therefore, adaptive bit and power loading methods are of great interest for PLC. In this work, we will evaluate a cyclic loading method for an OFDM system in the CENELEC A-band. Our analysis is based on noise measurements at three different locations over an observation time of 23 hours. Our contribution includes an OFDM system model comprising both additive noise and channel attenuation. Furthermore, we present a method for limiting the amplitude and power spectral density to comply with narrowband regulations. In conclusion, our proposed method achieves a lower bit error rate at all locations. The improvement reaches up to a 10 2 times lower bit error rate compared to non-adaptive carrier allocation at the same data rate.

In order to make convolutional neural networks (CNNs) usable on smaller or mobile devices, it is necessary to reduce the computing, energy and storage requirements of these networks. One can achieved this by a fixed-point quantization of weights and activations of a CNN, which are usually represented by 32-bit floating-point. In this paper, we present an adaption of convolutional and fully connected layers in order to obtain a high usage of the given value range of activations and weights. Therefore, we introduce scaling factors obtained by moving average to limit the weights and activations. Our model, quantized to 8 bit, outperforms the 7-layer baseline model from which it is derived and the naive quantization by several percentage points. Our method does not require any additional operations in the inference and both the weights and activations have a fixed radix point.

Inspektionsprozesse werden im Remanufacturing auch heute noch vorwiegend manuell durchgeführt, da die Einschätzung des Qualitätszustands von rückläufigen Gebrauchtprodukten komplex und damit schwer zu automatisieren ist. Dies ist darauf zurückzuführen, dass Abnutzungsgrade, Deformationen und Schädigungen eine individuelle Bewertung des Gebrauchtprodukts nach sich ziehen und somit schwer standardisierbar sind. In diesem Beitrag werden die Anforderungen an ein System für die Bewältigung der Herausforderung der automatisierten Inspektion im Remanufacturing abgeleitet. Darauf aufbauend wird das Konzept einer Befundungsstation, welches diese Anforderungen erfüllt, präsentiert und Anwendungsfälle im Rahmen des von der Carl-Zeiss-Stiftung geförderten Forschungsprojekts „AgiProbot - Agiles Produktionssystem mittels mobiler, lernender Roboter mit Multisensorik bei ungewissen Produktspezifikationen“ vorgestellt.

No abstract

No abstract