Author´s reply

The author appreciates very much the comments on his work and the commenter\´s supplement of other important references related to the subject of system order reduction. With respect to the comment on the physical basis of the criterion he describes, the author maintains he has clearly stated in his article that the criterion is based on the philosophy that the low order model L(s) is chosen to approximate the frequency response of the higher order system H(s). On the point of using a vector criterion instead of a scalar criterion, there is no doubt that the former is, in theory, more powerful than the latter. However, the former may easily lead to optimization problems which are more difficult to solve. The author fully agrees with the commentor\´s view that the "best model" is highly dependent on the specific application involved. Therefore, a more logical way for obtaining a specific model is to incorporate into the simplification criterion, a priori, elements measuring the performance of the resulting overall system.