Research Papers - Dept of Software Engineering

Permanent URI for this collectionhttps://rda.sliit.lk/handle/123456789/1022

Browse

Filters

2019 - 201912020 - 20221

Advanced Search

Filter by

Settings

Author: search.filters.author.Jayawardena, C

Search Results

Now showing 1 - 2 of 2
  • Thumbnail Image
    PublicationEmbargo
    Train a Robot to Climb Staircase Using Vision-Base System
    (Institute of Electrical and Electronics Engineers, 2022-09-16) Jayawardana, J.T.H.; Dilshan, H.V.V; Wijethilaka, R.G.K.H; Balasooriya, T.D; Rajapaksha, U.U.S; Harshanath, S.M.B.; Jayawardena, C
    Currently, robots are used for different types of work, such as the manufacturing industry, healthcare, and the hotel industry. According to the current epidemic situation, the usage of robots was increased because of the need to reduce human interaction. As a result, they have to walk around the workplace, because of that, they may have to climb staircases. The world has many types of robots. Here the selected robot is a humanoid. This proposal is concerned with how to detect the staircase, count steps, get dimensions of it, and move the robot on it by keeping body balance. First, want to know what the objects are, then walk. The identified images from the image sensors will get as input. The technology stack that is used for image analysis is a method related to computer vision in deep learning. Other than that, while climbing the stairs robot needs to identify whether the staircase is over or not. Here we introduce a new concept: get the number of steps required to climb before climbing the stairs. It is related to how humans identify things by seeing and making decisions. The need to take the dimensions of the stairs is that when considering the stairs, they have their height, width, and range. Therefore, it is imperative to calculate the dimensions of each staircase separately. Creating a balance system similar to human balance is a great advantage in robotics. To implement such a system, there is an upright pose controller to allow the robot to walk stably by preventing tilting of the robot during walking on uneven floor. In general, for us to do that we need to calculate the global inclination of the floor is a key factor. It can be measured with a 2-axis accelerator meter, and it is installed in the inertial sensor.
  • Thumbnail Image
    PublicationEmbargo
    Comparative analysis of the application of Deep Learning techniques for Forex Rate prediction
    (IEEE, 2019-12-05) Aryal, S; Nadarajah, D; Kasthurirathna, D; Rupasinghe, L; Jayawardena, C
    Forecasting the financial time series is an extensive field of study. Even though the econometric models, traditional machine learning models, artificial neural networks and deep learning models have been used to predict the financial time series, deep learning models have been recently employed to do predictions of financial time series. In this paper, three different deep learning models called Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN) and Temporal Convolution Network (TCN) have been used to predict the United States Dollar (USD) to Sri Lankan Rupees (LKR) exchange rate and compared the accuracy of the models. The results indicate the superiority of CNN model over other models. We conclude that CNN based models perform best in financial time series prediction.