More information
Description
The research area developed by dr Olena Pavliuk is complementary to research scope of CoBotAGV project (NOR / POLNOR / CoBotAGV / 0027/2019 -00). Dr Olena Pavliuk has experience in machine learning methods including application of neural networks for energy consumption prediction. She has previous research experiences related to application of Artificial Intelligence for computer systems used for manufacturing. In the scope of CoBotAGV project she declared research on the use of neural networks to predict energy consumption by AGV (Autonomous Guided Vehicles) that will extend te work package
Summary of the results
PhD Olena Pavliuk has focused on the prediction of AGV (Autonomous Guided Vehicle ) energy consumption which is one of enabling methodology for AGVs’ task scheduling in real scenarios (unknown and uncertain environments). The nine deep-learning architectures and eight CWT configurations were tested, which resulted in 60 possible combinations and 300 models being programmed and tested. The best results were produced by the model with DenseNet121 architecture, Morlet wavelet and scale value from 0 to 256. A novel deep-learning model pre-trained on CWT-generated scalograms was proposed. The proposed pre-trained model was then tested on the pre-processed UCI-HAPT dataset and its subset to determine how the pre-trained model performs on target datasets of different sizes and with some significant distinctions from the source dataset. PhD Olena Pavliuk presented her research results by: (i) one publication submitted to Electronics journal (as the first author); (ii) two conference publications on IEEE BigData 2022 (as the first author/rank B in CORE DB) and IEEE CSIT; (iii) additionally, PhD Olena Pavliuk developed research on human activity recognition as an introduction to the area of analysis of cooperation between CoBotAGV and human staff. This part of research was presented as one journal and one conference publications (as the first author) on 5th International Conference on Informatics and Data-Driven Medicine.