Alpaslan Fığlalı

Investigation of Ant System parameter interactions by using design of experiments for job-shop scheduling problems

In recent years, one of the most important and promising research fields has been metaheuristics to find optimal or near-optimal solutions for NP-hard combinatorial optimization problems. Improving the quality of the solution or the solution time is basic research area on metaheuristics. Modifications of the existing ones or creation of hybrid approaches are the focus of these efforts. Another area of improving the solution quality of metaheuristics is finding the optimal combination of algorithm control parameters. This is usually done by design of experiments or one-at-a-time approach in genetic algorithms, simulated annealing and similar metaheuristics. We observe that, in studies which use Ant Colonies Optimization (ACO) as an optimization technique; the levels of control parameters are determined by some non-systematic initial experiments and the interactions of the parameters are not studied yet. In this study, the parameters of Ant System have been investigated on different sized and randomly generated job-shop scheduling problems by using design of experiments. The effects and interactions of the parameters have been interpreted with the outputs of the experiments. Referring to the statistical analysis it is observed that none of the interactions between the Ant System parameters has a significant effect on makespan value. A specific fractional experimental design is suggested instead of the full factorial design. Depending on the findings from the benchmark problems it will be a reliable approach to use the suggested design for saving time and effort in experiments without sacrificing the solution quality.

Fair referee assignments for professional football leagues

Assignment of referees to football games is an important problem faced in professional football leagues. Despite its importance, the problem has received limited academic attention. This paper presents a model and analysis of the problem for fair referee assignments, and develops a constructive heuristic and a local search procedure for its solution. Results from an extensive computational study show that the methods are effective in solving the problem in a second of computation time and yielding an excellent solution quality.

Image processing-aided working posture analysis

OWAS method is adapted to an integrated software as a prototype.It operates completely computer-aided with the help of image processing techniques.That models performance is high while robust recording conditions can be settled.Necessity of expert analyst is eliminated.The model will support the common use of OWAS in industry. Musculoskeletal Disorders (MSDs) rank among the commonest health problems both in the frequency of concurrency and in the money spent on these disorders, which mainly stem from poor working posture it also negatively affects employees in terms of job productivity, life quality, and both physical and social activities. Analyzing and improving working postures with scientific methods provides significant contributions in the field of controlling job performance and decreasing MSDs. OWAS (Ovako Working Posture Analyzing System) is one of the methods for analyzing working postures and can be applied to very diverse areas successfully. In this study, a prototype of integrated software, which is based on image processing techniques, was developed (I-OWAS), and the performance of the model was presented. I-OWAS begins with separating the video film into frames, producing OWAS codes belonging to working posture in each frame, and then classifying the images according to risk categories. Despite OWAS being a successful method for analyzing working postures, it requires an expert analysis. Also the manual analyzing process is so laborious and time consuming. I-OWAS provide the computer support for the manual coding stage and eliminates the need for an expert analyst; hence, the method can be widely used in industry.

Artificial neural networks for the cost estimation of stamping dies

Abstract Although artificial neural networks (ANN) are more known in the field of image recognition and forecasting, cost estimation has become another emerging area in recent years. In this study, the establishment of an intelligent system was attempted for forecasting the total cost of sheet metal stamping dies. In this context, where the cost of stamping dies is estimated with a conventional approach which has been applied in the company up to now, the ANN and multiple regression analysis and the performance of the three cost-estimation models are examined. The examinations are based on the data of previous costs and use a number of critical criteria which are decided by experienced tool makers and engineers from every level of the organization of the seven companies which produce stamping dies. The comparative study reveals that the ANN system outperforms the traditional linear regression analysis model and conventional approach used for cost estimation. Thus, it is possible for firms which produce stamping dies to obtain a fairly accurate prediction with an ANN model and determined criteria.

Feasibility analysis of a hydraulic ram pump investment project

Usage and management of water resources issues are considered world’s environmental problems of primary importance. Half of the global workforce is employed in several water and natural resource-dependent industries, i.e., agriculture, energy, building, and transport. Due to these reasons, rationally planning and usage of water resources is essential. The most common use of water in the world is in the agriculture sector with the rate of 70%. Since water resources are one of the most important inputs for agriculture, the use of this resource in agricultural irrigation with the right methods has great importance in terms of sustainability of agricultural production. Considering that a great percentage of agricultural areas are very far from the electricity supply due to the geographical conditions, the hydraulic ram pump, taking advantage of basic physic and mechanic principles, the wide spread usage of hydraulic ram pump is crucial. Our attention has been given to the fact that in the abovementioned conditions, a feasibility study of a hydraulic ram investment project was carried out with the proper methods of watering. Hydraulic ram is an environmentally friendly water pumping device with no fuel and no electricity consumption. Two scenarios were examined in the feasibility study. In the first scenario, a feasibility study was carried out for the leasing of a workshop to manufacture hydraulic rams, while all required procedures for manufacturing a water coach were outsourced in the second scenario. As a result of the study, these two evaluated scenarios were presented and interpreted.

Yapay Sinir Ağı Yaklaşımıyla Çinko Kalınlığının Tahminlenmesi

Since the competition level among the companies is increasing day by day, meeting customer demands with qualified products and cost reduction are primary goals of each company. And zinc, the main raw material in galvanization sector, is the most important cost item. So it is required to forecast the amount of zinc to be spent. In this study it is tried to forecast the amount of zinc consumption using the artificial neural network (ANN) method. To evaluate the convenience of values hypothesis tests are done; and the results showed that there is no significant difference between the predicted and real outputs statistically

A Model to Increase Customer Loyalty by Using Bi-directional Semantic Interference: An Application to White Goods Industry

Adding values on products, services, or systems by responding to customers’ needs quickly is a relevant issue for companies and requires them to be located competitively in their environment. In line with this issue, the demand chain approach enables companies to create user-centered designs. A demand chain is defined as a special customer-oriented supply chain network structure in the decision making process, that analyses customer demand and market conditions in order to reach an efficient distribution. The aim of this study is to examine the customer portfolio in white goods industry and find the preferences of current and potential customers in order to build and retain customer loyalty. The decision rules which help to increase the market share are obtained by using the Classification and Regression Trees (CART). After conducting a comprehensive literature survey on the demand chain approach and its applications, the required components for a network structure are determined by utilizing the bi-directional relationships between the customers and the manufacturers. Hence, eighty-five customer and twenty dealer surveys are carried out. The research methodology is then presented and the data collected from the surveys is analyzed statistically. The results of the study have prominent importance to overcome the uncertainty in demand chain and to determine which strategies should be adopted by the companies to have a loyal customer base.