It is very important to establish the control model for deformation path during the process of metal forming. Using the techniques of Artificial Neural Network(ANN) and numerical simulation, a new approach controlling deformation path based on the identification of material parameter is proposed. In this approach, the identification of material parameter value m and the control of deformation path proceed one by one in each loading stage. The identification of material parameter is finished by ANN using the experimental strain increments, loading increments and material stresses, after that, we can get loading increments for the next deformation stage by the ANN trained by material stresses, objective strain increments and the identified value m. By tension and torsion test of thin walled tube on MTS testing machine, the experimental strain path was obtained and compared with objective one. The results of this comparison validate the proposed approach for the deformation path control.