Decomposed piecewise-linear models by hyperplanes unbending

This paper presents a mathematical procedure to obtain decomposed piecewise-linear (PWL) models. The models have been denoted as decomposed because they include the independent and dependent variables separately in a same formulation. The mathematical procedure has been denoted as hyperplanes unbending. It has been applied into the explicit model of Chua and the implicit model of Bokhoven resulting a recasting of these models into decomposed forms which can be applied to achieve the representation of univalued and multivalued PWL characteristics. The hyperplanes unbending technique is specifically useful in the task of finding a model description for PWL characteristics which are doubly-multivalued because it implies decomposing the original function, that describes the doubly-multivalued PWL characteristic, in two univalued functions related by a mapping variable