FUZZY LOGIC AND FUZZY ARTIFICIAL NEURAL NETWORK USE IN ENGINEERING DESIGN PROCESS

Rapid development of production methods with technology has shortened product life cycle of products. This makes it necessary for the design process of new products to be faster and less costly. In this thesis, classical techniques and soft-computing techniques that accelerate engineering design processes have been investigated. It is examined at which stages of the engineering design process these techniques were used. In addition, two different soft-computing techniques were developed for two different engineering design processes and used in a foundry industry. In the first application, fuzzy logic is applied in the process of gathering information which is a sub process of engineering design process. In this application, fuzzy logic model has been formed in selection of cast irons produced according to certain standards. The results of the fuzzy logic application and the results of the model were compared with the results of the experts. In the second application, fuzzy artificial neural network model is developed in parametric design process which is the sub process of engineering design process. Fuzzy artificial neural network model is used to determine the effect of casting parameters on hardness output. Each parameter is translated into triangular fuzzy numbers and entered the artificial neural network. The results of the generated model were statistically tested whether the system represents the parameters. It is thought that this thesis will contribute to the research of alternative methods in accelerating engineering design processes in the casting sector despite the studies.