An Overview of Statistical Prediction Models for Solar Energy Harvesting Wireless Sensor Networks
Sohail SARANG, Goran M STOJANOVIĆ, Stevan STANKOVSKI, Varun JEOTI
In conventional wireless sensor networks (WSNs), nodes are powered by small batteries which have a limited lifetime. The energy harvesting techniques enable the development of energy harvesting WSNs (EH-WSNs) that can harvest power from the surrounding environment using different energy sources to achieve sustainable node operation and perform different tasks. However, the energy harvested from these sources depends significantly on the surrounding conditions that exhibit significant variations. Consequently, uncertainty in the harvestable energy hampers nodes to plan their energy resources more effectively and incorporate smart energy allocation strategies to improve network performance. Therefore, forecasting the amount of harvestable energy from external sources will enable nodes to utilize the energy resources accordingly and consider high variability of energy intake to avoid any power outages in the near future. Thus, several prediction models have been proposed for EH-WSNs, specifically for solar energy to predict harvestable energy. In this paper, existing statistical prediction models have been described thoroughly in terms of statistical approach and prediction accuracy.
Branislav DUDIĆ, Pavel KOVAČ, Borislav SAVKOVIĆ
To succeed in the world market, companies must invest in the purchase of technology such as industrial robots. Industrial robots have become very important tools that give manufacturers an edge over the competition in an increasingly challenging global market. In the production process, industrial robots are able to perform a given activity better than humans in terms of quality of work. Robots are able to work continuously 24 hours a day, whereas, for comparison, human working time is on average 8-10 hours. We also need to know that robots are flexible, so they are able to respond to a change in production faster than humans. Manufacturers who use this technology are gaining an effective economic tool that allows them to gain an advantage on the world market. Over the last 5 years, the global market has recorded a record in the number of robots delivered and manufactured, where the total robot market is around $ 45 billion. The countries that are the largest customers of robots are China, South Korea, Japan, the USA and Germany. The aim of this work is to point out the very important and significant role of industrial robots, to show the situation on the world market and what are the innovations and investments in this field, also to give guidelines on how the market will behave in the next period and which are the most important companies manufactured by industrial robots.
Modeling of Torque and Thrust Force in Drilling Using Genetic Algorithm
Milenko SEKULIĆ, Dragan RODIĆ, Marin GOSTIMIROVIĆ, Borislav SAVKOVIĆ, Anđelko ALEKSIĆ, Nenad KULUNDŽIĆ
Modeling the drilling process is very important as a tool for understanding the process and solving practical problems. This is especially important under the conditions of high cutting speeds and feeds associated with mass production. Mathematical drilling models are particularly important for determining thrust force and torque. A realistic estimation of these quantities is important both for the design of the drill geometry and for solving the main problems in production (surface defects, vibrations, tool wear, burr formation, etc.). Over the years, various mechanical-mathematical models have been developed to determine the cutting forces. In this paper, the application of a genetic algorithm for modeling torque and thrust force in the drilling operation is presented. The predicted thrust force and torque, based on the models of GA, show good agreement with the experimental values.
Slobodan TABAKOVIĆ, Saša ŽIVANOVIĆ, Zoran DIMIĆ, Milan ZELJKOVIĆ
The paper presents the programming and program verification on a 3-axis hybrid kinematics CNC machine for rapid prototyping using subtractive and additive processes. The original hybrid (parallel-serial) 3-axis O-X glide mechanism developed to build a rapid prototyping machine and multifunctional machine tools is presented. The paper analyzes the available programming software, which can be one of the standard CAD/CAM systems or a specialized CAM system, for subtractive processes, i.e. desktop milling. For the additive processes, the software for generating G code based on the STL file as well as the possibility of simulating the machine when working is considered. To verify the program, the simulation of material removal for subtractive processes as well as the simulation of material addition for additive processes were considered. The paper presents the prototype of a hybrid kinematics CNC machine and some of the results of testing with an open control system based on the LinuxCNC.