DYPIC - Dynamic Positioning in Ice: First Phase of Model Testing

2012 ◽  
Author(s):  
Andrea Haase ◽  
Solange van der Werff ◽  
Peter Jochmann
Keyword(s):  
Tracking System ◽  
Development Project ◽  
Model Testing ◽  
Fine Tuning ◽  
Model Parameters ◽  
Second Phase ◽  
Dynamic Positioning ◽  
The European Union ◽  
Testing Procedures ◽  
Ice Conditions

DYPIC (Dynamic Positioning in Ice) is a research and development project within the MARTEC ERA-NET project of the European Union. Its objective is to contribute to the closure of the gap between DP in open water being an industry standard, and DP in ice which has some extra challenges to tackle. Two phases of model testing in ice form the back bone of the project and are facilitated by HSVA (Hamburg Ship Model Basin, Germany). The first test phase, which was executed from May to July 2011, involved two different model ships. Both were tested in free floating mode (where the model sailed solely by its own propulsion system) and fixed mode (where the model was connected to a carriage). In the free floating mode the controlling was performed by a prototype DP system scaled to model parameters. Four different managed ice fields with systematically varied ice concentration and ice floe size were prepared in the ice tank in order to investigate the influence of the relevant parameters. Tests were executed for several velocities and headings with respect to the approaching ice floes. In the free floating case ice loads on the hull were derived from the measured loads on the thrusters. The behavior of the model ship was captured by the position and heading tracking system Qualisys and several installed video cameras. The fixed mode tests serve well as a reference measurement. The results will be used to develop a model scale DP system for ice that is adjustable to different kinds of vessels and ice conditions and eventually to develop testing procedures for the assessment of the DP performance of a vessel in managed ice. A second phase of model testing for fine tuning and benchmarking the developed system will be carried out in August 2012. Within the scope of the paper is the description of the performed tests speaking of test setup and ice conditions. Analyses of results are not covered.

DYPIC - Dynamic Positioning in Ice: Second Phase of Model Testing

2013 ◽  
Author(s):  
Andrea Haase ◽  
Peter Jochmann
Keyword(s):  
Phase I ◽  
Phase Ii ◽  
Development Project ◽  
Model Testing ◽  
Second Phase ◽  
Dynamic Positioning ◽  
Station Keeping ◽  
Level Ice ◽  
Ice Field ◽  
Ice Model

DYPIC - Dynamic Positioning in Ice is a European research and development project where the main goal is to customize a dynamic positioning (DP) system for model testing in an ice model basin. To achieve this objective numerous ice model tests are performed. Overall they are divided into two main phases — DYPIC Phase I in 2011 and DYPIC Phase II in 2012. The first phase is documented and presented in [1]. This paper addresses the description of the second phase and the presentation of a selection of results. As the main goal of Phase II is to test the DP system developed in Phase I the trials of the second phase are mainly performed in DP mode, while very few tests that serve separate sub goals within the project are performed in the so called fixed mode where the model is towed through the tank. For the DP mode different configurations of the test setup itself are tested. In order to simulate station keeping the vessel travels either in front or behind the main carriage trying to hold its position relatively to the carriage. The relative motion is captured by optical cameras on the carriage and markers on the vessel. In addition real station keeping tests are performed while the model stayed in the middle of the ice basin and different ice field types are pushed along. The ice features tested in DYPIC Phase II include managed ice fields of different kinds and level ice.

Influence of the Ice Concentration on the Ice Loads on the Hull of a Ship in a Managed Ice Field

2012 ◽  
Author(s):  
Solange van der Werff ◽  
Andrea Haase ◽  
René Huijsmans ◽  
Qin Zhang
Keyword(s):  
Development Project ◽  
Model Tests ◽  
Ice Thickness ◽  
Model Parameters ◽  
Dynamic Positioning ◽  
Ship Model ◽  
Ice Concentration ◽  
Ice Loads ◽  
Ice Field ◽  
Yaw Angle

The research and development project DYPIC (Dynamic Positioning in Ice) focuses on the challenges related to DP operations in Arctic environment. At the HSVA (Hamburg Ship Model Basin, Germany), model tests in ice were carried out using two configurations; one where the model was fixed to the towing carriage, and a free floating configuration, where the model ship was controlled by a DP system scaled to model parameters. During the model tests a number of parameters were systematically varied. Model ship velocity and yaw angle were the parameters related to the controlling of the model. In addition, the ice field characteristics were varied by applying two variations in ice floe size and two variations in concentration, resulting in four different ice field descriptions. The ice thickness was remained constant for all test implementations. Every test run with a particular controlling (velocity and heading) profile was executed in each of the four ice fields. In order to develop a DP controller which is optimally adjusted to the environment in which the system operates, it is important to find relations between the characteristics of the ice field and the forces they apply on the hull of the vessel or construction. An assessment of the measurements and observations during the testing is the basis of a study which has the objective to find how the ice field appearance and the ice loads on a structure relate to each other.

PO-1739 Optimal threshold of model parameters for the respiratory tracking system with helical tomotherapy

2021 ◽  
Vol 161 ◽  
pp. S1461-S1462
Author(s):  
W. Okada ◽  
M. Tanooka ◽  
H. Doi ◽  
K. Sano ◽  
M. Shibata ◽  
...  
Keyword(s):  
Helical Tomotherapy ◽  
Tracking System ◽  
Optimal Threshold ◽  
Model Parameters ◽  
Respiratory Tracking

scholarly journals Self-Supervised Pre-Training of Transformers for Satellite Image Time Series Classification

2020 ◽  
Author(s):  
Yuan Yuan ◽  
Lei Lin
Keyword(s):  
Time Series ◽  
Deep Learning ◽  
Large Scale ◽  
Temporal Structure ◽  
Satellite Image ◽  
Fine Tuning ◽  
Small Scale ◽  
Model Parameters ◽  
Learning Approaches ◽  
Wide Range

Satellite image time series (SITS) classification is a major research topic in remote sensing and is relevant for a wide range of applications. Deep learning approaches have been commonly employed for SITS classification and have provided state-of-the-art performance. However, deep learning methods suffer from overfitting when labeled data is scarce. To address this problem, we propose a novel self-supervised pre-training scheme to initialize a Transformer-based network by utilizing large-scale unlabeled data. In detail, the model is asked to predict randomly contaminated observations given an entire time series of a pixel. The main idea of our proposal is to leverage the inherent temporal structure of satellite time series to learn general-purpose spectral-temporal representations related to land cover semantics. Once pre-training is completed, the pre-trained network can be further adapted to various SITS classification tasks by fine-tuning all the model parameters on small-scale task-related labeled data. In this way, the general knowledge and representations about SITS can be transferred to a label-scarce task, thereby improving the generalization performance of the model as well as reducing the risk of overfitting. Comprehensive experiments have been carried out on three benchmark datasets over large study areas. Experimental results demonstrate the effectiveness of the proposed method, leading to a classification accuracy increment up to 1.91% to 6.69%. <div><b>This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible.</b></div>

scholarly journals Analysis of Impact of Ship Model Parameters on Changes of Control Quality Index in Ship Dynamic Positioning System

2019 ◽  
Vol 26 (1) ◽  
pp. 6-14 ◽  
Author(s):  
Tacjana Niksa Rynkiewicz ◽  
Anna Witkowska
Keyword(s):  
Quality Index ◽  
Adaptive Controller ◽  
Model Parameters ◽  
Dynamic Positioning ◽  
Positioning System ◽  
Integral Control ◽  
Dynamic Positioning System ◽  
Ship Model ◽  
Control Quality ◽  
The Impact

Abstract In this work there is presented an analysis of impact of ship model parameters on changes of control quality index in a ship dynamic positioning system designed with the use of a backstepping adaptive controller. Assessment of the impact of ship model parameters was performed on the basis of Pareto-Lorentz curves and ABC method in order to determine sets of the parameters which have either crucial, moderate or low impact on objective function. Simulation investigations were carried out with taking into account integral control quality indices.

scholarly journals Unsupervised Learning in RSS-Based DFLT Using an EM Algorithm

Sensors ◽  
2021 ◽  
Vol 21 (16) ◽  
pp. 5549
Author(s):  
Ossi Kaltiokallio ◽  
Roland Hostettler ◽  
Hüseyin Yiğitler ◽  
Mikko Valkama
Keyword(s):  
Em Algorithm ◽  
Expectation Maximization ◽  
Tracking System ◽  
Model Parameters ◽  
Calibration Data ◽  
Tracking Accuracy ◽  
Time Period ◽  
The Em Algorithm ◽  
Device Free ◽  
Localization And Tracking

Received signal strength (RSS) changes of static wireless nodes can be used for device-free localization and tracking (DFLT). Most RSS-based DFLT systems require access to calibration data, either RSS measurements from a time period when the area was not occupied by people, or measurements while a person stands in known locations. Such calibration periods can be very expensive in terms of time and effort, making system deployment and maintenance challenging. This paper develops an Expectation-Maximization (EM) algorithm based on Gaussian smoothing for estimating the unknown RSS model parameters, liberating the system from supervised training and calibration periods. To fully use the EM algorithm’s potential, a novel localization-and-tracking system is presented to estimate a target’s arbitrary trajectory. To demonstrate the effectiveness of the proposed approach, it is shown that: (i) the system requires no calibration period; (ii) the EM algorithm improves the accuracy of existing DFLT methods; (iii) it is computationally very efficient; and (iv) the system outperforms a state-of-the-art adaptive DFLT system in terms of tracking accuracy.

scholarly journals Navigating Through the Maze: Practical Constraints in the Implementation of Forest Regulations (FLEGT) in Ghana

2021 ◽  
Vol 10 (2) ◽  
pp. 31
Author(s):  
Gordon Kofi Sarfo-Adu
Keyword(s):  
Value Chain ◽  
Sustainable Forest Management ◽  
Tracking System ◽  
Action Plan ◽  
The European Union ◽  
Implementation Challenges ◽  
Technical Errors ◽  
Public Policy Implementation ◽  
Transnational Actors ◽  
The Eu

The European Union Forest Law Enforcement on Governance and Trade (EU-FLEGT) Action Plan seeks to promote widespread sustainable forest management and relies largely on transnational actors and international law in its operationalization. The EU FLEGT sets out EU custom regulation through Voluntary Partnership Agreements (VPAs) which is a bilateral agreement between the EU and wood exporting countries with instruments aimed at promoting sustainable practices within the forest resources value chain. Ghana became a signatory to the FLEGT VPA since 2007, as part of the process, it is required to use technology to track timber logging from source to point of export. Issues of networks and inter-agency collaboration and dealing with human elements remain crucial in ensuring effective operationalization. Adopting a qualitative case study design as well as theories and concepts from the public policy implementation literature, this study examines the implementation vagaries of the FLEGT VPA in Ghana. Although the VPA is a laudable idea of using Information Technology (IT) in effectively tracking timber to its original source to ascertain legality or otherwise of the timber, the needed IT infrastructure and resources have not matched up with the goal. Additionally, the VPA implementation is expensive and has come with additional cost to the implementers, The study further observes that the increasing ‘red flags’ that are raised on the Ghana Wood Tracking System is a blend of technical errors emanating from negligence or capacity challenges and human manipulation. This calls for regular consultations and workshops with relevant stakeholders in order to assess which skills are deficient and a need to beef up through on-the-job training. The domestic market and trading activities tend to fuel demand for illegal timber hence a constraint to the full realization of the VPA objective. The study makes policy suggestions on how to address these implementation challenges.

scholarly journals Learning to Learn Morphological Inflection for Resource-Poor Languages

2020 ◽  
Vol 34 (05) ◽  
pp. 8058-8065
Author(s):  
Katharina Kann ◽  
Samuel R. Bowman ◽  
Kyunghyun Cho
Keyword(s):  
Fine Tuning ◽  
Target Language ◽  
Model Parameters ◽  
Transfer Model ◽  
Absolute Accuracy ◽  
Learning To Learn ◽  
Suggested Approach ◽  
High Resource ◽  
Resource Poor ◽  
Cross Lingual

We propose to cast the task of morphological inflection—mapping a lemma to an indicated inflected form—for resource-poor languages as a meta-learning problem. Treating each language as a separate task, we use data from high-resource source languages to learn a set of model parameters that can serve as a strong initialization point for fine-tuning on a resource-poor target language. Experiments with two model architectures on 29 target languages from 3 families show that our suggested approach outperforms all baselines. In particular, it obtains a 31.7% higher absolute accuracy than a previously proposed cross-lingual transfer model and outperforms the previous state of the art by 1.7% absolute accuracy on average over languages.

Utilizing a Thermal Time Model to Estimate Safe Times to Transplant Tilia Trees

2019 ◽  
Vol 45 (5) ◽  
Author(s):  
Tapio Linkosalo ◽  
Pilvi Siljamo ◽  
Anu Riikonen ◽  
Frank Chmielewski ◽  
Juha Raisio
Keyword(s):  
Geographical Area ◽  
Frost Damage ◽  
Thermal Time ◽  
Weather Data ◽  
Model Parameters ◽  
The European Union ◽  
Time Model ◽  
Large Size ◽  
Nursery Trees ◽  
Thermal Time Model

City trees planted in parks and along streets are typically grown to large size in nurseries before being transplanted to their final growing sites. According to tendering rules within the European Union (EU), any business may compete for public contracts in any EU country, and this applies to purchases of valuable lots of nursery trees. There is however a risk of poor transplanting success if the trees are imported from very distant locations with a different pace of spring development. The aim of this study was to implement a Thermal Time model to predict the spring development of Tilia trees to find out in which geographical area the spring development is sufficiently similar to conditions in southern Finland, so that the success of transplantation of the trees is not unduly risked. We used phenological observations collected at the International Phenological Gardens (IPGs) over the whole of Europe, together with ERA-Interim weather data to estimate the model parameters, and then used the same date to predict the onset of leaf unfolding ofTilia during the years 1980 to 2015. Producing maps of phenological development of Tilia, we concluded that there are no large risks of frost damage if tree import area is limited to northern parts of Baltics or to the west coast of Scandinavia.

Colorisation of archival aerial imagery using deep learning

2021 ◽  
Author(s):  
Ryusei Ishii ◽  
Patrice Carbonneau ◽  
Hitoshi Miyamoto
Keyword(s):  
Deep Learning ◽  
Urban Expansion ◽  
Research Work ◽  
Fine Tuning ◽  
Aerial Images ◽  
Model Parameters ◽  
Style Transfer ◽  
Freeze Layer ◽  
Rgb Images ◽  
Japanese Rivers

&lt;p&gt;Archival imagery dating back to the mid-twentieth century holds information that pre-dates urban expansion and the worst impacts of climate change.&amp;#160; In this research, we examine deep learning colorisation methods applied to historical aerial images in Japan.&amp;#160; Specifically, we attempt to colorize monochrome images of river basins by applying the method of Neural Style Transfer (NST).&amp;#160;&amp;#160;&amp;#160; First, we created RGB orthomosaics (1m) for reaches of 3 Japanese rivers, the Kurobe, Ishikari, and Kinu rivers.&amp;#160; From the orthomosaics, we extract 60 thousand image tiles of `100 x100` pixels in order to train the CNN used in NST.&amp;#160; The Image tiles were classified into 6 classes: urban, river, forest, tree, grass, and paddy field.&amp;#160; Second, we use the VGG16 model pre-trained on ImageNet data in a transfer learning approach where we freeze a variable number of layers.&amp;#160; We fine-tuned the training epochs, learning rate, and frozen layers in VGG16 in order to derive the optimal CNN used in NST.&amp;#160; The fine tuning resulted in the F-measure accuracy of 0.961, 0.947, and 0.917 for the freeze layer in 7,11,15, respectively.&amp;#160; Third, we colorize monochrome aerial images by the NST with the retrained model weights.&amp;#160; Here used RGB images for 7 Japanese rivers and the corresponding grayscale versions to evaluate the present NST colorization performance.&amp;#160; The RMSE between the RGB and resultant colorized images showed the best performance with the model parameters of lower content layer (6), shallower freeze layer (7), and larger style/content weighting ratio (1.0 x10&amp;#8309;).&amp;#160; The NST hyperparameter analysis indicated that the colorized images became rougher when the content layer selected deeper in the VGG model.&amp;#160; This is because the deeper the layer, the more features were extracted from the original image.&amp;#160; It was also confirmed that the Kurobe and Ishikari rivers indicated higher accuracy in colorisation.&amp;#160; It might come from the fact that the training dataset of the fine tuning was extracted from these river images.&amp;#160; Finally, we colorized historical monochrome images of Kurobe river with the best NST parameters, resulting in quality high enough compared with the RGB images.&amp;#160; The result indicated that the fine tuning of the NST model could achieve high performance to proceed further land cover classification in future research work.&lt;/p&gt;

Sign in / Sign up

Export Citation Format

Share Document