Study of Threshing Function of Combine harvester with Artificial Neural Network

Taguchi Method, one of methods of robust engineering, has been a leading tool in thequality because S/N ratio as in this method represents factorial effect of control variables to outputsconsidering noise factors. According to the standard procedure of Taguchi, it is needed to haveexperiments depending on Lattice matrix such as L18 matrix. On the other hand, artificial neuralnetwork (ANN) has been utilized to describe a lot of systems because it is capable of describing nonlinearsystem. In this paper, threshing function of combine harvester was modeled by ANN. Trainingof ANN was done with the experimental results not depending on Lattice matrix. After training, ANNmodel was used as an additional tool to Taguchi Method in design of threshing function of combineharvester in order to calculate the factorial effect of control variables. It was proved that proposedmethod combining Taguchi Method and ANN is effective.