An Efficient Hybrid Approach for Performance-Based Alternative Design Evaluation in Systems Engineering

In this research, an innovative method to evaluate the efficiency of different designs in system engineering has been presented. This method is based on the combination of two strong techniques, namely DEMATEL and artificial neural networks. By employing the DEMATEL method, relationships between different design features are established. Subsequently, the DEMATEL matrix is converted and fed into an artificial neural network for training. To train the network, high-quality training data has been used and the performance of the model has been evaluated using various evaluation criteria such as mean squared error and coefficient of determination. The results show that the proposed model performs very well in the evaluation of alternative designs and is significantly superior to traditional methods. Also, by performing sensitivity analysis, the effect of each feature on the performance of the model has been investigated. Finally, in this research, suggestions for further improvement and development of this method have been presented. This research can be used as a powerful tool for decision-making in the process of designing and selecting engineering plans.