Wyniki wyszukiwania - BazTech

1

Neuroevolutionary approach to COLREGs ship maneuvers

Łącki M.

TransNav : International Journal on Marine Navigation and Safety of Sea Transportation

|

2019

|

Vol. 13. no. 4

745--750

EN

The paper describes the usage of neuroevolutionary method in collision avoidance of two power-driven vessels approaching each other regarding COLREGs rules. This may be also be seen as the ship handling system that simulates a learning process of a group of artificial helmsmen - autonomous control units, created with artificial neural networks. The helmsman observes an environment by its input signals and according to assigned CORLEGs rule, he calculates the values of required parameters of maneuvers (propellers rpm and rudder deflection) in a collision avoidance situation. In neuroevolution such units are treated as individuals in population of artificial neural networks, which through environmental sensing and evolutionary algorithms learn to perform given task safely and efficiently. The main task of this project is to evolve a population of helmsmen which is able to effectively implement chosen rule: crossing or overtaking.

2

Analysis of Methods of Determining the Safe Ship Trajectory

Lisowski J.

TransNav : International Journal on Marine Navigation and Safety of Sea Transportation

|

2016

|

Vol. 10, no. 2

223--228

EN

The paper describes six methods of optimal and game theory and artificial neural network for synthesis of safe control in collision situations at sea. The application of optimal and game control algorithms to determine the own ship safe trajectory during the passing of other encountered ships in good and restricted visibility at sea is presented. The comparison of the safe ship control in collision situation: multi-step matrix non-cooperative and cooperative games, multi-stage positional non-cooperative and cooperative games have been introduced. The considerations have been illustrated with examples of computer simulation of the algorithms to determine safe of own ship trajectories in a navigational situation during passing of eight met ships.

3

Consistently Trained Artificial Neural Network for Automatic Ship Berthing Control

Ahmed Y., Hasegawa K.

TransNav : International Journal on Marine Navigation and Safety of Sea Transportation

|

2015

|

Vol. 9, no. 3

417--426

EN

In this paper, consistently trained Artificial Neural Network controller for automatic ship berthing is discussed. Minimum time course changing manoeuvre is utilised to ensure such consistency and a new concept named ‘virtual window’ is introduced. Such consistent teaching data are then used to train two separate multi-layered feed forward neural networks for command rudder and propeller revolution output. After proper training, several known and unknown conditions are tested to judge the effectiveness of the proposed controller using Monte Carlo simulations. After getting acceptable percentages of success, the trained networks are implemented for the free running experiment system to judge the network’s real time response for Esso Osaka 3-m model ship. The network’s behaviour during such experiments is also investigated for possible effect of initial conditions as well as wind disturbances. Moreover, since the final goal point of the proposed controller is set at some distance from the actual pier to ensure safety, therefore a study on automatic tug assistance is also discussed for the final alignment of the ship with actual pier.

4

Using topological data analysis for diagnosis pulmonary embolism

Rucco M., Merelli E., Herman D., Ramanan D., Petrossian T., Falsetti L., Nitti C., Salvi A.

Journal of Theoretical and Applied Computer Science

|

2015

|

Vol. 9, nr 1

41--55

EN

Pulmonary Embolism (PE) is a common and potentially lethal condition. Most patients die within the first few hours from the event. Despite diagnostic advances, delays and underdiagnosis in PE are common. Moreover, many investigations pursued in the suspect of PE result negative and no more than 10% of the pulmonary angio-CT scan performed to confirm PE confirm the suspected diagnosis. To increase the diagnostic performance in PE, current diagnostic work-up of patients with suspected acute pulmonary embolism usually starts with the assessment of clinical pretest probability using plasma d-Dimer measurement and clinical prediction rules. One of the most validated and widely used clinical decision rules are the Wells and Geneva Revised scores. However, both indices have limitations. We aimed to develop a new clinical prediction rule (CPR) for PE based on a new approach for features selection based on topological concepts and artificial neural network. Filter or wrapper methods for features reduction cannot be applied to our dataset: the application of these algorithms can only be performed on datasets without missing data. Alternatively, eliminating rows with null values in the dataset would reduce the sample size significantly and result in a covariance matrix that is singular. Instead, we applied Topological data analysis (TDA) to overcome the hurdle of processing datasets with null values missing data. A topological network was developed using the Ayasdi-Iris software (Ayasdi, Inc., Palo Alto). The PE patient topology identified two flares in the pathological group and hence two distinct clusters of PE patient populations. Additionally, the topological network detected several sub-groups among healthy patients that likely are affected with non-PE diseases. To be diagnosed properly even though they are not affected by PE, in a next study we will introduce also the survival curves for the patients. TDA was further utilized to identify key features which are best associated as diagnostic factors for PE and used this information to define the input space for a back-propagation artificial neural network (BP-ANN). It is shown that the area under curve (AUC) of BP-ANN is greater than the AUCs of the scores (Wells and revised Geneva) used among physicians. The results demonstrate topological data analysis and the BP-ANN, when used in combination, can produce better predictive models than Wells or revised Geneva scores system for the analyzed cohort. The new CPR can help physicians to predict the probability of PE.

5

“P” coordinator scheme and interaction prediction principle in hierarchical structure of ANN

Płaczek S.

Computer Applications in Electrical Engineering

|

2015

|

Vol. 13

319--329

EN

When implementing the hierarchical structure [4][5] of the learning algorithm of an Artificial Neural Network (ANN), two very important questions have to be solved. The first one is connected with the selection of the broad coordination principle. In [1], three different principles are described. They vary with regard to the degree of freedom for the first-level tasks. The second problem is connected with the coordinator structure or, in other words, the coordination algorithm. In the regulation theory, the process of finding the coordinator structure is known as the feedback principle. The simplest regulator structure (scheme) is known as the proportional regulator – “P” regulator. In the article, the regulator structure and its parameters are analysed as well as their impact on the learning process quality.

6

Resistance Prediction for Hard Chine Hulls in the Pre-Planing Regime

Radojcic D., Zgradic A., Kalajdzic M., Simic A.

Polish Maritime Research

|

2014

|

nr 2

9--26

EN

A mathematical representation of calm-water resistance for contemporary planing hull forms based on the USCG and TUNS Series is presented. Regression analysis and artificial neural network (ANN) techniques are used to establish, respectively, Simple and Complex mathematical models. For the Simple model, resistance is the dependent variable (actually R/Δ for standard displacement of Δ = 100000 lb), while the Froude number based on volume (FnV) and slenderness ration (L/V1/3) are the independent variables. In addition to these, Complex model’s independent variables are the length beam ratio (L/B), the position of longitudinal centre of gravity (LCG/L) and the deadrise angle (β). The speed range corresponding to FnV values between 0.6 and 3.5 is analyzed. The Simple model can be used in the concept design phases, while the Complex one might be used for various numerical towing tank performance predictions during all design phases, as appropriate.

7

Reinforcement Learning in Ship Handling

Łącki M.

TransNav : International Journal on Marine Navigation and Safety of Sea Transportation

|

2008

|

Vol. 2, no. 2

157--160

EN

This paper presents the idea of using machine learning techniques to simulate and demonstrate learning behaviour in ship manoeuvring. Simulated model of ship is treated as an agent, which through environmental sensing learns itself to navigate through restricted waters selecting an optimum trajectory. Learning phase of the task is to observe current state and choose one of the available actions. The agent gets positive reward for reaching destination and negative reward for hitting an obstacle. Few reinforcement learning algorithms are considered. Experimental results based on simulation program are presented for different layouts of possible routes within restricted area.

8

An Application of ANN to Automatic Ship Berthing Using Selective Controller

Im N., Seong-Jong Lee, Hyung-Do B.

TransNav : International Journal on Marine Navigation and Safety of Sea Transportation

|

2007

|

Vol. 1, no. 1

101--105

EN

This paper deals with ANN(Artificial Neural Networks) and its application to automatic ship berthing. As ship motions are expressed by a multi-term non-linear model, it is very difficult to find optimal methods for automatic ship berthing. When a ship makes its berthing operation, the ship’s inertia and slow motion make the ship approach to final berthing point with pre-determined navigation pattern. If the ship is out of the pre-determined navigation pattern, the berthing usually end in failure. It has been known that the automatic control for ship’s berthing cannot cope with various berthing situations such as various port shape and approaching directions. For these reasons, the study on automatic berthing using ANN usually have been carried out based on one port shape and predetermined approaching direction. In this paper, new algorithm with ANN controller was suggested to cope with these problems. Under newly suggested algorithm, the controller can select different weight on the link of neural networks according to various situations, so the ship can maintain stable berthing operation even in different situations. Numerical simulations are carried out with this control system to find its improvement.

9

Application of real time recurrent neural network for seismic event detection

Wiszniowski J.

Acta Geophysica Polonica

|

2000

|

Vol. 48, nr 1

1-26

EN

The paper presents the detection of earthquakes by the Real Time Recurrent Neural Network (RTRN). The model of the network, the method of teaching it, and the results of earthquakes detection recorded by seismic stations of the Institute of Geophysics, Polish Academy of Sciences, are described. In a typical Artificial Neural Network (ANN), output values depend on instantaneous input values only. The RTRN has recurrent connections between network elements. The network outputs also depend on preceding input values. Information about a whole seismogram may be then used to detect instantaneous features, such as spectrum, and, at the same time, to detect time changes of the signal, with no need of supplying the input with long windows or spectrograms. Therefore, this method is able to efficiently detect regional earthquakes and teleseismic events. The network was investigated for the artificially imposed noised data and the real data taken from the stations with a high level of noise. The RTRN is able to generalize the results of teaching during recognition of other earthquakes.