Infeasibility in automated test assembly models: A comparison study of different methods