1. |
- Mansouri, Sina Sharif, et al.
(författare)
-
Towards MAV Navigation in Underground Mine Using Deep Learning
- 2018
-
Ingår i: IEEE ROBIO 2018. - : IEEE. ; , s. 880-885
-
Konferensbidrag (refereegranskat)abstract
- The usage of Micro Aerial Vehicles (MAVs) is rapidly emerging in the mining industry to increase overall safety and productivity. However, the mine environment is especially challenging for the MAV's operation due to the lack of illumination, narrow passages, wind gusts, dust, and other factors that can affect the MAV's overall flying capability. This article presents a method to assist the navigation of MAVs by using a method from the field of Deep Learning (DL), while considering a low-cost platform without high-end sensor suits. The presented DL scheme can be further utilized as a supervised image classifier that has the ability to process the image frames from a single on-board camera and to provide mine tunnel wall collision prevention. The efficiency of the proposed scheme has been experimentally evaluated in two underground tunnel environments that were used for data collection, training, and corresponding testing under multiple flying scenarios with different cameras configurations and illuminations.
|
|
2. |
- Georgoulas, George, 1976-, et al.
(författare)
-
An Automatic Method for Condition Monitoring of Inverter Fed Induction Motors
- 2018
-
Ingår i: Proceedings. - : IEEE. ; , s. 1754-1760
-
Konferensbidrag (refereegranskat)abstract
- This paper proposes an automatic method, for monitoring inverter fed induction motors using external stray flux measurements. The method relies on the marginal power spectrum of the Synchrosqueezed Wavelet Transform for the feature extraction stage and on Principal Component Analysis for the reduction of the high dimensionality of the generated feature vector. For the next stage two approaches were tested: a) a fault detector based on a one-class classifier and b) a fault diagnosis module based on a multiclass classifier. Both of them achieve high accuracies when tested with measurements coming from an experimental set up able to simulate stator short circuits and bearing faults. An explanation of the performance is given by visual inspection of the projection of the feature vectors into a three-dimensional space.
|
|
3. |
- Georgoulas, George, 1976-, et al.
(författare)
-
Exploring the Detectability of Short-Circuit Faults in Inverter-Fed Induction Motors
- 2018
-
Ingår i: Proceedings IECON 2018. - : IEEE. ; , s. 5930-5935
-
Konferensbidrag (refereegranskat)abstract
- This paper explores the possibility of creating an automatic method for assessing the condition of induction motor circuits fed by inverters. The stator current and magnetic flux are processed in the frequency domain and a feature selection stage is employed to pinpoint the most informative components to further be fed to a classifier that performs the assessment of the motor circuit. The results are promising, indicating that short circuit detection as well as quantification is feasible using noninvasive techniques.
|
|
4. |
- Karvelis, Petros, et al.
(författare)
-
Topic recommendation using Doc2Vec
- 2018
-
Konferensbidrag (refereegranskat)abstract
- The ever-increasing number of electronic content stored in digital libraries requires a significant amount of effort in cataloguing and has led to self-deposit solutions where the authors submit and publish their own digital records. Even in self-deposit, going through the abstract and assigning subject terms or keywords is a time consuming and expensive process, yet crucial for the metadata quality of the record that affects retrieval. Therefore, an automatic, or even a semi-automatic process that can recommend topics for a new entry is of huge practical value. A system that can address that has to rely basically on two components, one component for efficiently representing the relevant information of the new document and one component for recommending an appropriate set of topics based on the representation of the previous stage. In this work, different candidate solutions for both components are investigated and compared. For the first stage both distributed Document to Vector (doc2vec) and conventional Bag of Words (BoW) components are employed, while for the latter two different transformation approaches from the field of multi-label classification are compared. For the comparison, a collection of Ph.D. abstracts (~19000 documents) from the MIT Libraries Dspace repository is used suggesting that different combinations can provide high quality solutions.
|
|
5. |
- Mansouri, Sina Sharif, et al.
(författare)
-
2D visual area coverage and path planning coupled with camera footprints
- 2018
-
Ingår i: Control Engineering Practice. - : Elsevier. - 0967-0661 .- 1873-6939. ; 75, s. 1-16
-
Tidskriftsartikel (refereegranskat)abstract
- Unmanned Aerial Vehicles (UAVs) equipped with visual sensors are widely used in area coverage missions. Guaranteeing full coverage coupled with camera footprint is one of the most challenging tasks, thus, in the presented novel approach a coverage path planner for the inspection of 2D areas is established, a 3 Degree of Freedom (DoF) camera movement is considered and the shortest path from the taking off to the landing station is generated, while covering the target area. The proposed scheme requires a priori information about the boundaries of the target area and generates the paths in an offline process. The efficacy and the overall performance of the proposed method has been experimentally evaluated in multiple indoor inspection experiments with convex and non convex areas. Furthermore, the image streams collected during the coverage tasks were post-processed using image stitching for obtaining a single overview of the covered scene.
|
|