|
|
Train Identification System at Railway Switches And Crossings Using Advanced Machine Learning Methods
Krč, Rostislav ; Vorel,, Jan (referee) ; Plášek, Otto (referee) ; Podroužek, Jan (advisor)
This doctoral thesis elaborates possibilities of automatic train type identification in railway S&C using accelerometer data. Current state-of-the-art was considered, including requirements stated by research projects such as S-Code, In2Track or Turnout 4.0. Conducted experiments considered different architectures of artificial neural networks (ANN) and statistically evaluated multiple use case scenarios. The resulting accuracy reached up to 89.2% for convolutional neural network (CNN), which was selected as a suitable baseline architecture for further experiments. High generalization capability was observed as models trained on data from one location were able to classify locomotive types in the other location. Further experiments evaluated the effect of signal filtering and denoising. Evaluation of allocated memory and processing time for pre-trained models proved feasibility for in-situ application with regard to hardware restrictions. Due to a limited amount of available accelerometer data, distribution grid power demand data were utilized for further refinement of the proposed CNN architecture. Deep multi-layer architecture with regularization techniques such as dropout or batch normalization provides state-of-the-art performance for time series classification problems. Class activation mapping (CAM) allowed an explanation of decisions made by the neural network. Presented results proved that train type identification directly in the S&C is possible. The CNN was selected as optimal architecture for this task due to high classification accuracy, automatic filtration, and pattern recognition capabilities, allowing for the incorporation of the end-to-end learning strategy. Moreover, direct on-site application of pre-trained models is feasible with respect to limitations of in-situ hardware. This thesis contributes to understanding the train type identification problem and provides a solid theoretical background for future research.
|
|
|
Footbridge across the Labe
Krč, Rostislav ; Romportl, Tomáš (referee) ; Stráský, Jiří (advisor)
The thesis focuses on design of prestressed concrete footbridge which passes pedestrian and bicycle path over the Elbe river near city of Celakovice in the Czech Republic. Three different bridge options were analyzed and for further development a cable-stayed footbridge was chosen. Its bridge deck is formed of a box girder supported by cables in its vertical axis and all cable stays are anchored into two concrete A-shaped pylons. This structure was analyzed in SCIA Engineer and assessment of serviceability limit states and ultimate limit states according to recent European standards (Eurocodes) was made. Both the global structural behavior and the local structural integrity of box cross-section were assessed as well as construction stages and cross-sections of pylons and cable stays. Assessments were performed in IDEA StatiCa combined with hand calculations. Eventually a dynamic response of structure was analyzed. Natural modes and frequencies were found and forced oscillation response was evaluated. The thesis includes technical drawings, construction process and visualization.
|
|
|
Prestressed road bridge in Staré Město
Krč, Rostislav ; Trenz, Jan (referee) ; Koláček, Jan (advisor)
The thesis focuses on design of superstructure of concrete bridge, which passes a roadway across the river Krupa in the town of Stare Mesto. It is a simple span structure with length of 19 meters. For further elaboration, post-tensioned concrete slab with trapezoidal-shaped cross section was chosen. Assessment of serviceability limit states and ultimate limit states according to recent European standards (Eurocodes) was made. Furthermore, a design of post-tension according to former Czech technical standards (CSN) was studied. The thesis includes technical drawings and visualization.
|
|
|
Train Identification System at Railway Switches And Crossings Using Advanced Machine Learning Methods
Krč, Rostislav ; Vorel,, Jan (referee) ; Plášek, Otto (referee) ; Podroužek, Jan (advisor)
This doctoral thesis elaborates possibilities of automatic train type identification in railway S&C using accelerometer data. Current state-of-the-art was considered, including requirements stated by research projects such as S-Code, In2Track or Turnout 4.0. Conducted experiments considered different architectures of artificial neural networks (ANN) and statistically evaluated multiple use case scenarios. The resulting accuracy reached up to 89.2% for convolutional neural network (CNN), which was selected as a suitable baseline architecture for further experiments. High generalization capability was observed as models trained on data from one location were able to classify locomotive types in the other location. Further experiments evaluated the effect of signal filtering and denoising. Evaluation of allocated memory and processing time for pre-trained models proved feasibility for in-situ application with regard to hardware restrictions. Due to a limited amount of available accelerometer data, distribution grid power demand data were utilized for further refinement of the proposed CNN architecture. Deep multi-layer architecture with regularization techniques such as dropout or batch normalization provides state-of-the-art performance for time series classification problems. Class activation mapping (CAM) allowed an explanation of decisions made by the neural network. Presented results proved that train type identification directly in the S&C is possible. The CNN was selected as optimal architecture for this task due to high classification accuracy, automatic filtration, and pattern recognition capabilities, allowing for the incorporation of the end-to-end learning strategy. Moreover, direct on-site application of pre-trained models is feasible with respect to limitations of in-situ hardware. This thesis contributes to understanding the train type identification problem and provides a solid theoretical background for future research.
|
|
|
Footbridge across the Labe
Krč, Rostislav ; Romportl, Tomáš (referee) ; Stráský, Jiří (advisor)
The thesis focuses on design of prestressed concrete footbridge which passes pedestrian and bicycle path over the Elbe river near city of Celakovice in the Czech Republic. Three different bridge options were analyzed and for further development a cable-stayed footbridge was chosen. Its bridge deck is formed of a box girder supported by cables in its vertical axis and all cable stays are anchored into two concrete A-shaped pylons. This structure was analyzed in SCIA Engineer and assessment of serviceability limit states and ultimate limit states according to recent European standards (Eurocodes) was made. Both the global structural behavior and the local structural integrity of box cross-section were assessed as well as construction stages and cross-sections of pylons and cable stays. Assessments were performed in IDEA StatiCa combined with hand calculations. Eventually a dynamic response of structure was analyzed. Natural modes and frequencies were found and forced oscillation response was evaluated. The thesis includes technical drawings, construction process and visualization.
|
|
|
Prestressed road bridge in Staré Město
Krč, Rostislav ; Trenz, Jan (referee) ; Koláček, Jan (advisor)
The thesis focuses on design of superstructure of concrete bridge, which passes a roadway across the river Krupa in the town of Stare Mesto. It is a simple span structure with length of 19 meters. For further elaboration, post-tensioned concrete slab with trapezoidal-shaped cross section was chosen. Assessment of serviceability limit states and ultimate limit states according to recent European standards (Eurocodes) was made. Furthermore, a design of post-tension according to former Czech technical standards (CSN) was studied. The thesis includes technical drawings and visualization.
|
guest ::
