Coefficients of constitutive relationship for metallic rubber were studied and trained by BP (back propagation) neural net.BP neural net was introduced elaborately

and how to predict coefficients of constitutive relationship by BP neural net was discussed.BP neural net parameters were chosen.Through parameter identification of static experimental data of metallic rubber

BP neural net models were obtained in two cases

including:various densities

various shape factors and densities;and further prediction for the nonlinear constitutive relationship was realized.The experiments prove that the theory agrees with the experiment

indicating that the method of predicting nonlinear constitutive relationship of metallic rubber is feasible.