The past decades have seen an increasing complexity of nonlinear material models proposed as scientific interests in materials grow and high performance materials are introduced. One of the core processes in the material modelling is the parameter identification, where constants in a material model are identified such that the model can describe material behaviour accurately. Due to the non-linearity of the material model, the parameter identification problem is most commonly converted into minimisation of an objective function describing errors between the measured and computed outputs using an optimisation method. However, the problem of this iterative technique is that the solution often cannot be obtained if measurement data and/or the model contain large errors [
] as this makes the objective function complex.