Volume 4, Issue 4, 2019.

Vision Guidance for Autonomous Vehicles: Comparative Study Of 3D Video and Point Cloud

Dragorad MILOVANOVIĆ, Dragan KUKOLJ, Sandra NEMET

Recent advances in 3D sensing and volumetric data compression are encouraging research efforts to meet the demands for vision guidance of autonomous vehicle. Nowadays, autonomous vehicle technology is successful in a high percentage of common road scenarios. However, new research efforts are required to meet the demands for higher performance. The diversity of traffic scenarios in the urban environment presents great challenges, foremost for video-based environment recognition. Autonomous driving technology and other automated assistance systems process huge amounts of data, thus efficient data compression, storage and retrieval is necessary. Key technologies for autonomous vehicles, i.e. 3D video/Point Cloud compression solutions for camera and LiDAR sensor systems are presented in this paper.

Feature Selection Using Combined Particle Swarm Optimization and Artificial Neural Network Approach

Sandra NEMET, Gordana OSTOJIĆ, Dragan KUKOLJ, Stevan STANKOVSKI, Dragan JOVANOVIĆ

This paper deals with identification of the most influencing input attributes related to the accuracy of the prediction model. It is assumed that the prediction model may be represented by any machine learning-based models, including artificial neural networks, fuzzy models, etc. Selection of influencing attributes is based on particle swarm optimization (PSO) combined with neural networks. The role of neural networks is to estimate fitness of each particle during the search procedure implemented using a PSO algorithm. The presented feature selection method represents the first step in the prediction model design. The method is applied on a dataset characterized with weak correlation between the model’s inputs and outputs. Selection of appropriate inputs improves the prediction model accuracy.

Modelling the Fuel Consumption in Hybrid and Electronic Airplanes using Matlab Aerospace toolbox

Sándor CSIKÓS,  László GOGOLÁK, Tamás MOLNÁR, Péter SZUCHY, István BÍRÓ, József SÁROSI

With the rise in popularity of electric cars more types of vehicles are adapting electric drives. In the case of airplanes the standards are higher than in the case of cars so the emphasis of research and innovation is greater. This paper presents a list of the challenges electric aircrafts face and their potential solutions. The paper describes the most important elements of electric and hybrid aircrafts. Comprehensive model was created taking into account the criteria set for hybrid and electric aircraft. With the aid of the model the effect of the most important components on efficiency can be assessed. One modelling approach in Matlab Simulink environment is also described. For the modelling the Matlab Aerospace toolbox is also used.

Simulation and Digital Prototyping – Valuable Inputs in the Optimization Process of an Advanced Prototype

Zoltan GOBOR

Based on the novel development methods while bearing in mind the state-of-the-art hardware and software solutions, a precision plot seeder for medicinal and aromatic plants was designed and built at the Bavarian State Research Center for Agriculture. The applied system-based design enabled the emphasis to be put on: simulation, validation and verification as well as identifying problems at the system level in the early phase of the design process. Multiple design variants were created to overcome the lack of the cross-disciplinary knowledge in order to consequently achieve savings in associated labor and material costs. Simplified analysis of circumstances under which the roughness of the soil surface negatively affects the maintenance of the predefined sowing depth by means of kinematical analysis is presented to demonstrate the advantages of using a digital model for development of new features.