Wyniki wyszukiwania - BazTech

1

Maximising accuracy and efficiency of traffic accident prediction combining information mining with computational intelligence approaches and decision trees

Tambouratzis T>, Souliou D., Chalikias M., Gregoriades A.

Journal of Artificial Intelligence and Soft Computing Research

|

2014

|

Vol. 4, No. 1

31--42

EN

The development of universal methodologies for the accurate, efficient, and timely prediction of traffic accident location and severity constitutes a crucial endeavour. In this piece of research, the best combinations of salient accident-related parameters and accurate accident severity prediction models are determined for the 2005 accident dataset brought together by the Republic of Cyprus Police. The optimal methodology involves: (a) information mining in the form of feature selection of the accident parameters that maximise prediction accuracy (implemented via scatter search), followed by feature extraction (implemented via principal component analysis) and selection of the minimal number of components that contain the salient information of the original parameters, which combined bring about an overall 74.42% reduction in the dataset dimensionality; (b) accident severity prediction via probabilistic neural networks and random forests, both of which independently accomplish over 96% correct prediction and a balanced proportion of under- and over-estimations of accident severity. An explanation of the superiority of the optimal combinations of parameters and models is given, as is a comparison with existing accident classification/prediction approaches.

2

Zastosowanie neuronowych klasyfikatorów grup podobieństwa dla wybranych wirtualnych obrazów wad materiałowych

Giżewski T., Wac-Włodarczyk A., Goleman R., Kowalski I.

Przegląd Elektrotechniczny

|

2011

|

R. 87, nr 12b

53-56

PL

W artykule przedstawiono metodologię klasyfikacji materiałów magnetycznych na podstawie charakterystycznego modelu statycznego, który jest analogiczny do funkcji gęstości klasycznego modelu Preisacha. Badana powierzchnia jest wirtualnym obrazem uzyskanym przez przetworzenie histerezy różnicowej. Jej postać umożliwia modelowanie rozkładu funkcji wagi, charakteryzującego dany typ wady materiałów ferromagnetycznych.

EN

The article presents a methodology for classification of magnetic materials based on the characteristic static model, which is analogous to the classical model of the density function Preisacha. The study area is a virtual image obtained by differential hysteresis processing. It’s character allows modeling of the weight function distribution, that is typical for the type of defects in ferromagnetic materials. (Neural classifier of similarity groups applied in selected virtual image of defects).

3

The Application of Neural Systems in Vibrodiagnosis

Romaniuk T.

Schedae Informaticae

|

2009

|

Vol. 17/18

21-44

EN

Vibrodiagnosis helps in detecting incipient faults in rotating machines like pumps and generators. Early detection prevents from undesired breakdown of the machine and allows to schedule maintenance times. The application of neural networks in classification of the rotating machine condition has been described in this work. Different types of networks and methods on feature extraction was described and compared. Additionally it was proposed a novelty feature set consisted of harmonics from vibration spectrum. The set were combined with using of probabilistic neural networks which has been modified that it could recognize defects that did not occur in the training set. Such architecture was tested in detection of two defects, shaft misalignment and mass unbalance. It was found that such network works better than a multi layered perceptron with statistical features.

4

Przewidywanie pęknięć w próbie SICO za pomocą probabilistycznych sieci neuronowych

Kusiak J., Godula A.

Hutnik, Wiadomości Hutnicze

|

2009

|

Vol. 76, nr 8

610-612

PL

W pracy podjęto próbę oceny możliwości i skuteczności probabilistycznej sztucznych sieci neuronowych w przewidywaniu pęknięć w procesach plastycznej przeróbki metali na gorąco. Zastosowano probabilistyczne sieci neuronowe (PNN – probabilistic neural networks), które zostały zaimplementowane do programu komercyjnego FORGE3. Weryfikację opracowanego modelu procesu pękania przeprowadzono na podstawie analizy wyników laboratoryjnej próby SICO. Wyniki przewidywania pękania porównano z wynikami uzyskanymi z kryterium pękania Lathama-Cockcrofta.

EN

The main goal of the work is an attempt of using the artificial neural networks in prediction of steel cracking during hot deformation. Developed model based on the probabilistic neural networks (PNN) is implemented into commercial finite element code FORGE3. The model can be applied to simulate the material cracking in a SICO plastometric test as well as in a wide range of real industrial processes. The obtained results based on the proposed approach are compared with the results obtained using the Latham-Cockcroft criterion.

5

Zastosowanie metod probabilistycznych do przewidywania pęknięć w próbie SICO

Śledzińska A., Kusiak J.

Prace Naukowe Politechniki Warszawskiej. Mechanika

|

2007

|

z. 216

13--18

PL

W pracy podjęto wykorzystania oraz ocenę możliwości i skuteczności metod probabilistycznych w przewidywaniu pęknięć w procesach plastycznej przeróbki metali na gorąco. Wykorzystano probabilistyczne sieci neuronowe (PNN – prababilistic neural networks) oraz naiwny klasyfikator bayerowski (NBC Naive Bayes Classifier). Weryfikację opracowanych modeli procesu pękania przeprowadzono w oparciu o analizę wyników próby SICO. Wyniki przewidywania pękania porównano z wynikami uzyskanymi z kryterium pękania Lathama-Cockrofta.

EN

Possibility of application of the probabilistic methods and to prediction of material failure during hot deformation is the main subject of the work. Developed model based on the probabilistic neural networks (PNN) and Naive Bayes Classifier (NBC) is implemented into commercial finite element code FORGE3. The model can be applied to simulate possibility of material failure in a simple plastometric tests (i.e. SICO test) and in a wide range of real industrial processes. The obtained results based on the proposed approach are compared with the results obtained using the Latham-Cockroft criterion, which is commonly used fracture criterion available in commercial FE software.