| 1. |
- Diez-Olivan, Alberto, et al.
(författare)
-
Adaptive Dendritic Cell-Deep Learning Approach for Industrial Prognosis Under Changing Conditions
- 2021
-
Ingår i: IEEE Transactions on Industrial Informatics. - : IEEE. - 1551-3203 .- 1941-0050. ; 17:11, s. 7760-7770
-
Tidskriftsartikel (refereegranskat)abstract
- Industrial prognosis refers to the prediction of failures of an industrial asset based on data collected by Internet of Things sensors. Prognostic models can experience the undesired effects of concept drift, namely, the presence of nonstationary phenomena that affects the data collected over time. Consequently, fault patterns learned from data become obsolete. To overcome this issue, contextual and operational changes must be detected and managed, triggering rapid model adaptation mechanisms. This article presents an adaptive learning approach based on a dendritic cell algorithm for drift detection and a deep neural network model that dynamically adapts to new operational conditions. A kernel density estimator with drift-based bandwidth is used to generate synthetic data for a faster adaptation, focusing on fine-tuning the lowest neural layers. Experimental results over a real-world industrial problem shed light on the outperforming behavior of the proposed approach when compared to other drift detectors and classification models.
|
|
| 2. |
- Diez-Olivan, Alberto, et al.
(författare)
-
Data Fusion and Machine Learning for Industrial Prognosis : Trends and Perspectives towards Industry 4.0
- 2018
-
Ingår i: Information Fusion. - : Elsevier. - 1566-2535 .- 1872-6305. ; 50, s. 92-111
-
Tidskriftsartikel (refereegranskat)abstract
- The so-called “smartization” of manufacturing industries has been conceived as the fourth industrial revolution or Industry 4.0, a paradigm shift propelled by the upsurge and progressive maturity of new Information and Communication Technologies (ICT) applied to industrial processes and products. From a data science perspective, this paradigm shift allows extracting relevant knowledge from monitored assets through the adoption of intelligent monitoring and data fusion strategies, as well as by the application of machine learning and optimization methods. One of the main goals of data science in this context is to effectively predict abnormal behaviors in industrial machinery, tools and processes so as to anticipate critical events and damage, eventually causing important economical losses and safety issues. In this context, data-driven prognosis is gradually gaining attention in different industrial sectors. This paper provides a comprehensive survey of the recent developments in data fusion and machine learning for industrial prognosis, placing an emphasis on the identification of research trends, niches of opportunity and unexplored challenges. To this end, a principled categorization of the utilized feature extraction techniques and machine learning methods will be provided on the basis of its intended purpose: analyze what caused the failure (descriptive), determine when the monitored asset will fail (predictive) or decide what to do so as to minimize its impact on the industry at hand (prescriptive). This threefold analysis, along with a discussion on its hardware and software implications, intends to serve as a stepping stone for future researchers and practitioners to join the community investigating on this vibrant field.
|
|
| 3. |
- Popescu, Adrian, et al.
(författare)
-
ENVIRAN : Energy Efficient Virtual Radio Access Networks
- 2013
-
Konferensbidrag (refereegranskat)abstract
- ENVIRAN is a new research project aiming at the research, design and deployment of new architectural solutions for network virtualization and cognitive radio networks. The project is about developing and testing a new network architecture, to enable innovation through programmability and control of network functions and protocols. For doing this, we solve different technical challenges. These are about network virtualization, open architecture, reconfigurable software suite, virtual base station and decision support system. Another important part of the project is regarding the development of a cognitive virtualization platform, to test the new developed solutions. It is well known that cognitive radio technology is a key concept suggested to use the radio frequency spectrum in a more efficient manner than previous mobile networks. The difference in our case is that the cognition is used not only to provide better resource use for bandwidth but also for other categories of resources like energy/power consumption (by using, e.g., green routing, cooperative/relay networking), hardware utilization (in form of, e.g., virtual Base Stations, cognitive/reconfigurable wireless devices), reduce the cost of supporting the required QoS/QoE, new business models. Cognition and virtualization concepts are used to increase the efficiency of network management and resource utilization as well as to reduce the power consumption and the cost of supporting the expected QoS/QoE for communication. The expected research results will be tested, among others, in the world-wide virtual network PlanetLab.
|
|
| 4. |
- Thymianis, Marios, et al.
(författare)
-
Electric Vehicle Routing Problem : Literature Review, Instances and Results with a Novel Ant Colony Optimization Method
- 2022
-
Ingår i: 2022 IEEE Congress on Evolutionary Computation, CEC 2022 - Conference Proceedings. - : IEEE. - 9781665467087
-
Konferensbidrag (refereegranskat)abstract
- One of the most well-known problems in combinatorial optimization is the Vehicle Routing Problem (VRP). Significant research has been done around this problem in two different perspectives: investigating new solving approaches, and studying variants of VRP which take into consideration multiple restrictions and constraints. One of such versions is the Electric Vehicle Routing Problem (EVRP), whose main objective is to find the optimal route of a fleet of electric vehicles, taking into account the locations of charging stations and the battery consumption of the mobile units. The aim of this study is threefold: (a) to perform a brief literature review on meta-heuristic approaches applied to the EVRP, (b) to offer insights on the available data instances for this problem, and (c) to discuss on the results of an experimental benchmark aimed at comparing different meta-heuristic approaches over diverse EVRP instances, including the proposal and evaluation of a novel Ant Colony Optimization approach.
|
|
