No abstract

This paper presents a novel method of improving energy efficiency of climbing robot by implementing accumulation of motion energy and its further recycling. Patented construction and functional scheme of pedipulator-conceptual walking mechanism are provided as well as a method to integrate movement and surface attachment modules of the robot. Calculations of dynamic loads are provided followed by results of simulations that demonstrate effective approach implementation by changing kinematic or constructional parameters instead of dynamic.

The object of research is preprint processes for the manufacture of audiobooks. The analytical and experimental studies reported here are based on the application of the methodology for devising a simulation model for calculating the complexity of preprint processes in the production of printed books with sound applications. The basic assumption of the study is that the use of simulation will contribute to the improvement of calculations of the complexity of preprint processes for making audiobooks. This cannot be achieved without analyzing the factors influencing the convenience of perception of textual and sound information and establishing the objective function of mutual influence of the text of the publication and audio. A methodology has been proposed for calculating the complexity of the preprint process of making audiobooks based on the developed algorithm of the technological process, mathematical representation of the structural model, as well as its components. It is shown that the process of creating audio accompaniment is affected by the content of printed pages, font design parameters, the selected format of page assembly, and the presence of illustrations. Based on the obtained expert assessments, a hierarchical representation of factors influencing the convenience of information perception is built; the interaction of parameters at different technological stages is described. The dependence of the complexity of the process of making audiobooks on the volume of the book edition in the accounting and publishing sheets, the size and capacity of the font, as well as the speed of sound recording of the printed text, has been established. It is determined that with an increase in font capacity within 200...1000 thousand characters, the calculated labor intensity increases by 5 % to 20 %. The largest limits of changes in the complexity of creating an audio accompaniment are 9 %, at the stage of processing graphic information – 50 %

This paper addresses the problem of insufficient data for text classification tasks and methods of datasets reinforcement. We explore the influence of data augmentation on classification accuracy. Text dataset is complemented by machine translations of similar data from languages of other families and the advantages of this approach are demonstrated. An experimental research is conducted for the sentiment analysis problem for hotel reviews data that shows the difference in accuracy of the model trained on original and translated data. Influence of dataset volume on resulting accuracy is studied and compared for both types of the datasets. It is shown that learning of the model graphs are similar and differs only with some minor static error.