Repository logo
 

Search Results

Now showing 1 - 5 of 5
  • Multiple-output switched-capacitor DC-DC combination converters for IoT applications
    Publication . Semiao, J.; Duran, E.; Litran, S. P.; Ferrera, M. B.
    This paper presents and describes switched-capacitor DC-DC combination converters with a single input and multiple outputs, with step-down and step-up voltage ratios. The combination converters provide two or three different DC outputs (depending on the combination), with different conversion ratios. The main advantage of the proposed converters is the low number of components used in the combination (two-output configurations require two link capacitors, and three-output configurations require three link capacitors) and two complementary phase clocks, which may reduce size and manufacturing costs for many applications. The DC-DC combination converters are designed to be fully integrated, and were implemented in a 65nm CMOS technology. The analysis of the proposed configurations for different performance modes and load conditions is presented, and simulation results show that these combination converter topologies are suitable for multiple-output switched-capacitor DC-DC converters with multiple conversion ratios.
  • Autonomous temporal pseudo-labeling for fish detection
    Publication . Veiga, Ricardo; Exposito Ochoa, Iñigo; Belackova, Adela; Bentes, Luis; Parente Silva, João; Semiao, J.; Rodrigues, João
    The first major step in training an object detection model to different classes from the available datasets is the gathering of meaningful and properly annotated data. This recurring task will determine the length of any project, and, more importantly, the quality of the resulting models. This obstacle is amplified when the data available for the new classes are scarce or incompatible, as in the case of fish detection in the open sea. This issue was tackled using a mixed and reversed approach: a network is initiated with a noisy dataset of the same species as our classes (fish), although in different scenarios and conditions (fish from Australian marine fauna), and we gathered the target footage (fish from Portuguese marine fauna; Atlantic Ocean) for the application without annotations. Using the temporal information of the detected objects and augmented techniques during later training, it was possible to generate highly accurate labels from our targeted footage. Furthermore, the data selection method retained the samples of each unique situation, filtering repetitive data, which would bias the training process. The obtained results validate the proposed method of automating the labeling processing, resorting directly to the final application as the source of training data. The presented method achieved a mean average precision of 93.11% on our own data, and 73.61% on unseen data, an increase of 24.65% and 25.53% over the baseline of the noisy dataset, respectively.
  • A Zeta-CSC converter combination for single-input and bipolar output
    Publication . Litran, S. P.; Duran, E.; Ferrera, M. B.; Semiao, J.; Barroso, R. S.
    This paper presents a DC-DC converter of one input and one bipolar output. The converter has been obtained from the combination of two basic converters of one input and one output. Thus, the proposed configuration is the result of the combination of a CSC (Canonical Switching Cell) converter and a Zeta converter. A common feature of these converters is they have the same conversion relation. The input stage formed by the switching device and an inductance is shared by both converters. On the other hand, the output of the CSC converter is inverted with respect to its input and the output of Zeta converter is non-inverted, so this combination results in a bipolar type output. With the proposed topology it is possible to reduce the number of components and eliminate the need for synchronization of several switching devices. The converter has been analyzed in a steady state and the most significant operating equations have been deduced. A simulation platform has been developed in the MATLAB-Simulink, which has allowed the behavior of the converter to be verified.
  • A distributed load scheduling mechanism for micro grids
    Publication . Monteiro, J. M.; Eduardo, J.; Cardoso, Pedro J. S.; Semiao, J.
    Several protocols have recently been defined for smart grids that enable the communication between electric devices and energy management systems. While these protocols and architectures can already be applied in different fields of micro grids, it is still not clear how the distributed resources and constraints of such electrical grids can be managed in an optimum way. In order to achieve a reduction in electricity costs and maximizing investments made in renewable sources, an optimization mechanism should be used to perform load scheduling, considering different variables such as forecasted power generation curve from renewable sources, different tariffs' rates, electric circuit constraints, user restrictions and correspondent comfort levels. Given these considerations, this work defines and evaluates a distributed micro grid resource management architecture and protocol which is able to optimize load scheduling while considering all the mentioned restrictions and parameters. The proposed architecture was implemented on a multi-agent simulator and the performed tests show that significant reductions in electricity cost can be achieved using this methodology. © 2014 IEEE.
  • Analysis of converters with bipolar output for DC microgrid
    Publication . Litran, S. P.; Duran, E.; Barroso, R. S.; Semiao, J.; Ferrera, M. B.
    Bipolar DC networks have been shown as a suitable solution for low voltage (LV) energy distribution. This paper presents the advantages and disadvantages that they have with respect to other topologies, from the point of view of their implementation, efficiency in the use of energy and its ability to connect large loads. On the other hand, for this type of networks the use of DC-DC converters is required to connect the different sources of distributed generation with guarantees so that the network as a whole can be managed automatically (Smartgrid), both in isolated mode and connected to the conventional AC network. In this line, three configurations of non-isolated DC-DC converters, of a bipolar input and output are presented. That allows connecting DC sources to the network in a controlled way. The behaviour of the converters is verified through a simulation platform based on MATLAB-Simulink and the experimental results will be presented.