Nursel Öztürk

Supplier selection and performance evaluation in just-in-time production environments

Research highlights? Neural network is used for supplier selection and evaluation in JIT production. ? The system can assist manufacturers in two ways: selecting and evaluating suppliers. ? The proposed systems are tested with data taken from an automotive factory. ? The results show that the proposed systems can be used effectively and simply. The purpose of this paper is to aid just-in-time (JIT) manufacturers in selecting the most appropriate suppliers and in evaluating supplier performance. Many manufacturers employ the JIT philosophy in order to be more competitive in todays global market. The success of JIT on the production floor has led many firms to expand the JIT philosophy to the entire supply chain. The procurement of parts and materials is a very important issue in the successful and effective implementation of JIT; thus, supplier selection and performance evaluation in long-term relationships have became more critical in JIT production environments. The proposed systems can assist manufacturers in handling these issues. In this research, neural network based supplier selection and supplier performance evaluation systems are presented. The proposed approach is not limited to JIT supply. It can assist manufacturers in selecting the most appropriate suppliers and in evaluating supplier performance. The proposed neural network based systems are tested with data taken from an automotive factory, and the results show that the proposed systems can be used effectively.

Neuro-Genetic Design Optimization Framework to Support the Integrated Robust Design Optimization Process in CE

This article describes an integrated and optimized product design framework to support the design optimization applications in concurrent engineering (CE). The significant consideration is given to show the effectiveness of hybrid approaches and how they can be used to improve the performance of integrated design optimization applications. The proposed approach is based on two-stages which are (1) the use of neural networks (NNs) and genetic algorithm (GA) with feature technology for integrated design activities and (2) the use of Taguchi’s method and GA for design parameters optimization. The first stage resulted in better integrated design solutions in terms of computational complexity and later resulted in a solution, which leads to better and more robust parameter values for multi-objective shape design optimization. The effectiveness and validity of the proposed approach are evaluated with examples.

Hybrid neural network and genetic algorithm based machining feature recognition

In this research, neural networks (NNs) and genetic algorithms (GAs) are used together in a hybrid approach to reduce the computational complexity of feature recognition problem. The proposed approach combines the characteristics of evolutionary technique and NN to overcome the shortcomings of feature recognition problem. Consideration is given to reduce the computational complexity of network with specific interest to design the optimum network architecture using GA input selection approach. In order to evaluate the performance of the proposed system, experimental results are compared with previous NN based feature recognition research.

Neural network based non-standard feature recognition to integrate CAD and CAM

Abstract In this paper, a neural network based feature recognition approach which is capable of extracting information from design database is proposed to automate the integration of the design and applications following design. CAD data base is converted to feature based model information which can be used by CAM applications. Multilayer perceptron neural network is provided with Boundary representation (B-rep) information to recognise simple and complex features. B-rep structure is used to process the face-score values in terms of geometry and topology of the solid model. The effectiveness of proposed approach is demonstrated with experimental results which show the validity of this method to recognise complex shape features.

Use of genetic algorithm to design optimal neural network structure

In this research, neural network (NN) and genetic algorithm (GA) are used together to design optimal NN structure. The proposed approach combines the characteristics of GA and NN to reduce the computational complexity of artificial intelligence applications in design and manufacturing. Genetic input selection approach is introduced to obtain optimal NN topology. Experimental results are given to evaluate the performance of the proposed system.

Hybrid approach for genetic algorithm and Taguchi's method based design optimization in the automotive industry

Although genetic algorithm and multi-objective optimization techniques are widely used to solve problems in the design and manufacturing area, further improvements are required to develop more efficient techniques regarding multi-objective optimization problems. The main goal of the present research is to further develop and strengthen the genetic algorithm based multi-objective optimization approach to generate real-world design solutions in the automotive industry. In this research, a new hybrid approach based on Taguchis method and a genetic algorithm is presented to achieve better Pareto-optimal set solutions for multi-objective design optimization problems. In addition, fatigue damage and life are also considered to evaluate the results of the design optimization process. The validity and efficiency of the proposed approach are evaluated and illustrated with test problems taken from the literature. It is then applied to a vehicle component taken from the automotive industry.

An advanced hybrid meta-heuristic algorithm for the vehicle routing problem with backhauls and time windows

A new model is built up for vehicle routing problem with backhauls and time windows.A hybrid algorithm which includes tabu search and simulated annealing is proposed.The nearest neighbor method is improved for initial solution generation.Proposed hybrid meta-heuristic algorithm outperforms the existing methods.34 new best solutions are obtained for 45 instances. This paper presents an advanced hybrid meta-heuristic algorithm (HMA) to solve the vehicle routing problem with backhauls and time windows (VRPBTW). The VRPBTW is an extension of the vehicle routing problem with time windows (VRPTW) and the vehicle routing problem with backhauls (VRPB) that includes capacity, backhaul and time window constraints. In this problem, the customers are divided into two subsets consisting of linehaul and backhaul customers. Each vehicle starts from the depot, and goods are delivered from the depot to the linehaul customers. Goods are subsequently returned to the depot from the backhaul customers. The objective is to minimize the total distance that satisfies all of the constraints. The proposed meta-heuristic method is tested on a problem data set obtained from Solomons VRPTW benchmark problems which includes 25, 50 and 100 demand nodes. The results of the computational studies show that the HMA outperforms the existing studies and provides better solutions than the best known solutions in practical computational times.

A memory structure adapted simulated annealing algorithm for a green vehicle routing problem

Currently, reduction of carbon dioxide (CO2) emissions and fuel consumption has become a critical environmental problem and has attracted the attention of both academia and the industrial sector. Government regulations and customer demands are making environmental responsibility an increasingly important factor in overall supply chain operations. Within these operations, transportation has the most hazardous effects on the environment, i.e., CO2 emissions, fuel consumption, noise and toxic effects on the ecosystem. This study aims to construct vehicle routes with time windows that minimize the total fuel consumption and CO2 emissions. The green vehicle routing problem with time windows (G-VRPTW) is formulated using a mixed integer linear programming model. A memory structure adapted simulated annealing (MSA-SA) meta-heuristic algorithm is constructed due to the high complexity of the proposed problem and long solution times for practical applications. The proposed models are integrated with a fuel consumption and CO2 emissions calculation algorithm that considers the vehicle technical specifications, vehicle load, and transportation distance in a green supply chain environment. The proposed models are validated using well-known instances with different numbers of customers. The computational results indicate that the MSA-SA heuristic is capable of obtaining good G-VRPTW solutions within a reasonable amount of time by providing reductions in fuel consumption and CO2 emissions.

Simulated annealing approach in scheduling of virtual cellular manufacturing in the automotive industry

An approach based on simulated annealing algorithm and heuristic method is presented as an efficient means of scheduling the manufacturing operations of virtual cellular manufacturing systems in the automotive industry. The objectives are to minimise the total weighted tardiness of the production schedule and to minimise the total materials travelling distance. A two-stage approach is employed for scheduling the manufacturing operations. In the first stage, the simulated annealing algorithm is applied to get the optimal schedule. In the second stage, a heuristic approach that was presented by Mak et al. (2007) is employed with some adjustments to minimise the total materials travelling distance. Examples are introduced to evaluate the performance of the present approach and to illustrate how the approach is employed to tackle scheduling problems. The results show that the approach is quite successful and can be used for scheduling the virtual manufacturing cells for the production of parts in case of frequently changing demands.

Computationally Efficient Approach for the Integration of Design and Manufacturing in CE

Today, companies are faced with fierce competition which is characterized by the necessity to bring the higher quality prod ucts and lower priced products to the market in shorter times than their competitors. The key to the success of organizations is the effec tive integration of design and applications following design such as machining, process planning, analysis, assembly, inspection etc. It was seen that effectiveness of the traditional CIM systems is not satisfactory to ensure competitiveness and high productivity. Recently, the concept of CE has been proposed to overcome the problems exist in integration. However, it has been recognized by both academic and industrial environments that efficient application of CE is still not achieved. In this research, STEP based feature recognition using neural networks is presented to develop feature based model and to enhance the integration of production activities in CE.