Bingol, SedatKihcgedik, Hidir Yanki2024-04-242024-04-2420180941-06431433-3058https://doi.org/10.1007/s00521-016-2718-5https://hdl.handle.net/11468/14498Design of the die in hot metal forming operations depends on the required forming load. There are several approaches in the literature for load prediction. Artificial neural networks (ANNs) have been successfully used by a few researches to estimate the forming loads. This paper aims at using the effectiveness of a new evolutionary approach called gene expression programming (GEP) for the estimation of forging load in hot upsetting and hot extrusion processes. Several parameters such as angle (alpha), L/D ratio (R), friction coefficient (A mu), velocity (v) and temperature (T) were used as input parameters. The accuracy of the developed GEP models was also compared with ANN models. This comparison was evidenced by some statistical measurements (R (2), RMSE, MAE). The outcomes of the study showed that GEP can be used as an effective tool for representing the complex relationship between the input and output parameters of hot metal forming processes.eninfo:eu-repo/semantics/closedAccessMetal FormingGepAnnUpsettingExtrusionForgingArticle303937945WOS:0004385954000202-s2.0-8500071483310.1007/s00521-016-2718-5Q1Q1