INESC-ID

About: INESC-ID is a nonprofit organization based out in Lisbon, Portugal. It is known for research contribution in the topics: Computer science & Context (language use). The organization has 932 authors who have published 2618 publications receiving 37658 citations.

Mapping the Impact of Daytime and Overnight Electric Vehicle Charging on Distribution Grids

Abstract: Strong adoption dynamics of private passenger electric vehicles (EV) will require adjustments in the operation and planning of electrical distribution grids. This work proposes a novel approach to assess the impact of electric vehicle charging while considering EV adoption dynamics and commuting patterns. The proposed model uses Geographic Information Systems (GIS) and is applied to a real-world case study. Results suggest that clustering of EV charging will occur and underline the relevance of accurate spatial and temporal charging pattern estimations for distribution grid planning. Overloading of distribution network elements was observed even under light EV penetration rates.

Large-Scale Peer-to-Peer Autonomic Monitoring

Abstract: The increasing scale and complexity of distributed system motivates the need for autonomous management. One of the key aspects in the management of distributed systems is the issue of component monitoring. Component monitoring is particularly challenging in large-scale dynamic systems, given the need to ensure that each component is monitored by at least one non-faulty component, despite joins, leaves, and failures, both at node and at network level. This paper proposes that components self-organize in an unstructured overlay network of constant degree in order to ensure that each component is always monitored by a threshold of other components.

Toward Better Chinese Word Segmentation for SMT via Bilingual Constraints

Abstract: This study investigates on building a better Chinese word segmentation model for statistical machine translation. It aims at leveraging word boundary information, automatically learned by bilingual character-based alignments, to induce a preferable segmentation model. We propose dealing with the induced word boundaries as soft constraints to bias the continuous learning of a supervised CRFs model, trained by the treebank data (labeled), on the bilingual data (unlabeled). The induced word boundary information is encoded as a graph propagation constraint. The constrained model induction is accomplished by using posterior regularization algorithm. The experiments on a Chinese-to-English machine translation task reveal that the proposed model can bring positive segmentation effects to translation quality.

Fingerprinting for Web Applications: From Devices to Related Groups

Abstract: Identifying users and user devices is as important in web applications as in many other contexts. In web applications, user identification usually involves an authentication process, e.g., providing a username and a password. Identification is also possible without explicit authentication using cookies or device fingerprints. Device fingerprinting is also useful for other purposes, e.g., to serve as a second factor of authentication. Recently some interest appeared in the problem of cross-device fingerprinting, i.e., of the identification of the same user in different devices using fingerprinting. We target a variation of the problem that we call related group fingerprinting. We define a related group as a set of persons (e.g., a family) that share the same home network. We devised a related group fingerprinting scheme that we evaluated experimentally with data from hundreds of users. This evaluation suggests that group fingerprinting is feasible.

Digital LQR control with Kalman Estimator for DC-DC Buck converter

Abstract: Energy management is critical in electronic systems and an adequate energy transfer is fundamental. Recent circuit requirements lead us to solutions such as the digital controllers, which can accommodate the classical control methods and new algorithms impractical in analog circuits. The linear-quadratic regulators (LQR) offer interesting properties that can be used to control a DC-DC Buck Converter. LQR, when applied with a Kalman State Estimator, can implement a control method that only senses the output voltage, avoiding a current sensor. In this work, a low sampling frequency Digital LQR with a Digital Kalman State Estimator with Buck Converter, working at 2 MHz, is implemented in Matlab. Classical control methods, such as Proportional-Integral-Derivative (PID) with High-Frequency Pole, Average Current-Mode with external Voltage compensation loop, Analog LQR with Kalman State Estimator and the Digital LQR are implemented in Matlab for comparison purposes.