The Experts below are selected from a list of 9 Experts worldwide ranked by ideXlab platform
Makhtoumi Mehra - One of the best experts on this subject based on the ideXlab platform.
-
Active Vibration Control of Launch Vehicle on Satellite Using Piezoelectric Stack Actuator
2018Co-Authors: Makhtoumi MehraAbstract:Satellites are subject to various severe vibration during different phases of flight. The concept of satellite smart Adapter is proposed in this study to achieve active vibration control of launch vehicle on satellite. The satellite smart Adapter has 18 active struts in which the middle section of each strut is made of piezoelectric stack actuator. Comprehensive conceptual design of the satellite smart Adapter is presented to indicate the design parameters, requirements and philosophy applied which are based on the reliability and durability criterions to ensure successful functionality of the proposed system, also fabrication process of the proposed piezoelectric stack actuator is discussed in detail. The coupled electromechanical virtual work equation for the piezoelectric stack actuator in each active strut is derived by applying D'Alembert's principle to calculate the consistent mass matrix, the stiffness matrix and the load vector using finite element approximation. Modal analysis is performed to characterize the inherent properties of the smart Adapter and extraction of a mathematical model of the system. Active vibration control analysis was conducted using fuzzy logic control with triangular membership functions and acceleration feedback. The control results conclude that the proposed satellite smart Adapter Configuration which benefits from piezoelectric stack actuator as elements of its 18 active struts has high strength and shows excellent robustness and effectiveness in vibration suppression of launch vehicle on satellite.Comment: 11 pages, 14 figures, 3 tables, 14 equation
Guo Xiao - One of the best experts on this subject based on the ideXlab platform.
-
design of multi net access Adapter Configuration and management based on embedded web
Journal of Military Communications Technology, 2007Co-Authors: Guo XiaoAbstract:Multi-net Access Adapter(MAA)is a transmit-control equipment used to interconnect long-distance local area networks.A design of multi-net access Adapter Configuration and management based on embedded web was presented.The main research contains forming an operating interface,constructing an embedded web server and developing an application program.
Goran Glavas - One of the best experts on this subject based on the ideXlab platform.
-
orthogonal language and task Adapters in zero shot cross lingual transfer
arXiv: Computation and Language, 2020Co-Authors: Marko Vidoni, Ivan Vulic, Goran GlavasAbstract:Adapter modules, additional trainable parameters that enable efficient fine-tuning of pretrained transformers, have recently been used for language specialization of multilingual transformers, improving downstream zero-shot cross-lingual transfer. In this work, we propose orthogonal language and task Adapters (dubbed orthoAdapters) for cross-lingual transfer. They are trained to encode language- and task-specific information that is complementary (i.e., orthogonal) to the knowledge already stored in the pretrained transformer's parameters. Our zero-shot cross-lingual transfer experiments, involving three tasks (POS-tagging, NER, NLI) and a set of 10 diverse languages, 1) point to the usefulness of orthoAdapters in cross-lingual transfer, especially for the most complex NLI task, but also 2) indicate that the optimal Adapter Configuration highly depends on the task and the target language. We hope that our work will motivate a wider investigation of usefulness of orthogonality constraints in language- and task-specific fine-tuning of pretrained transformers.
Marko Vidoni - One of the best experts on this subject based on the ideXlab platform.
-
orthogonal language and task Adapters in zero shot cross lingual transfer
arXiv: Computation and Language, 2020Co-Authors: Marko Vidoni, Ivan Vulic, Goran GlavasAbstract:Adapter modules, additional trainable parameters that enable efficient fine-tuning of pretrained transformers, have recently been used for language specialization of multilingual transformers, improving downstream zero-shot cross-lingual transfer. In this work, we propose orthogonal language and task Adapters (dubbed orthoAdapters) for cross-lingual transfer. They are trained to encode language- and task-specific information that is complementary (i.e., orthogonal) to the knowledge already stored in the pretrained transformer's parameters. Our zero-shot cross-lingual transfer experiments, involving three tasks (POS-tagging, NER, NLI) and a set of 10 diverse languages, 1) point to the usefulness of orthoAdapters in cross-lingual transfer, especially for the most complex NLI task, but also 2) indicate that the optimal Adapter Configuration highly depends on the task and the target language. We hope that our work will motivate a wider investigation of usefulness of orthogonality constraints in language- and task-specific fine-tuning of pretrained transformers.
Ivan Vulic - One of the best experts on this subject based on the ideXlab platform.
-
orthogonal language and task Adapters in zero shot cross lingual transfer
arXiv: Computation and Language, 2020Co-Authors: Marko Vidoni, Ivan Vulic, Goran GlavasAbstract:Adapter modules, additional trainable parameters that enable efficient fine-tuning of pretrained transformers, have recently been used for language specialization of multilingual transformers, improving downstream zero-shot cross-lingual transfer. In this work, we propose orthogonal language and task Adapters (dubbed orthoAdapters) for cross-lingual transfer. They are trained to encode language- and task-specific information that is complementary (i.e., orthogonal) to the knowledge already stored in the pretrained transformer's parameters. Our zero-shot cross-lingual transfer experiments, involving three tasks (POS-tagging, NER, NLI) and a set of 10 diverse languages, 1) point to the usefulness of orthoAdapters in cross-lingual transfer, especially for the most complex NLI task, but also 2) indicate that the optimal Adapter Configuration highly depends on the task and the target language. We hope that our work will motivate a wider investigation of usefulness of orthogonality constraints in language- and task-specific fine-tuning of pretrained transformers.