Ergonomic conditions of workplace settings is important for the performance of companies. Especially in the manufacturing industry, the employees are required to have convenient workplace conditions. If this is not the case, it is most likely to have a decrease in work efficiency, increase in workload, and negative impacts on employee health. In this study, we evaluate two ergonomic conditions, illumination and noise level, in different departments of a metal processing plant, to find the initial department to work on the improvement of ergonomic conditions. The evaluation of ergonomic conditions is done through a fuzzification process. The quantitative measurement results of illumination and noise level are fuzzified by Mamdani method. The fuzzified measurement values are scored with respect to specified interval lengths. As a result of this scoring process, ergonomically the worst conditioned department is found to start the improvement process.

No abstract

No abstract

Today flexible manufacturing systems are highly popular due to their capability of quick response to customer needs. Although the advantages of flexible manufacturing systems cannot be denied, these systems also bring new issues on production planning side. Especially assigning machines to production operations and scheduling these operations with respect to machine constraints turn out to be an NP-Hard problem. In this study, the integrated process routing and scheduling problem is explained, and the performance of two different meta-heuristic techniques, which are genetic algorithms and simulated annealing, are compared in terms of solution time and quality.

Increasing energy demand along with decreasing environmental resources necessitates looking for alternative energy sources. With this respect, solar power has gained considerable importance over recent years. This study analyzes the forecasting problem for the amount of electric power generated by solar power plants. The amount of electric power generated by solar power plants is not constant and changes depending on several variables such as the weather conditions, seasonal effects, t ype of solar panel, etc. On the other hand, to meet the electric power demand and minimize electric transfer cost, forecasting the electric power generated by solar power plants is critical. We test several neural network models with various weather-related input parameters. Among these parameters, we choose the most promising ones (radiation, humidity, hour, month) for further analysis to forecast the electric power generated by solar power plants located in the Konya region. Our test results over the past data show that it is possible to forecast the electric power generated by solar panels in the Konya region with less than 5% error.