The characteristics of pneumatic artificial muscles -or McKibben muscles -make them very interesting for the development of robotic applications that feature low impedance in terms of their interaction with the environment, such as can be the case with orthoses or certain wearable robots. In order to research the applicability of these actuators in industrial applications, an experimental one-degree-of-freedom set-up based on pneumatic muscles manufactured by Festo has been built at Ikerlan. This paper presents the modelling of a pneumatic muscle in Modelica as a new component. After that, the paper describes the set-up constructed and shows the complete model in Dymola/Modelica, in addition to validation of the model with some experimental data.

This paper deals with the design of stable adaptive control schemes for discrete plants under non-linear bounded inputs. Two algorithms are studied, one of them being a generalization of the d-step-ahead standard algorithm, while the second one involves the use of a time-decreasing adaptation gain. The class of inputs for which the algorithms perform in a stable fashion include saturations, relays and saturated dead-zones. The plant stability assumption is not requested to prove standard (non-asymptotic) stability results. The mechanism to achieve scheme's stability is the use of an extra weighted additive term in the adaptation error which takes into account the output deviation with respect to that associated with linear inputs. Nomenclature y(t), y*(t), .9(t) {z(t), t~O} UL(t), u(t) e. (I) e(t) e2(t) e(t) cP(t), cP;'(t) 0, O(t), 8'(t) O;"O;'(t) plant output, reference model output and estimated output through the estimated parameters, respectively z-sequence over the discrete set t = 0, 1, ... linear and non-linear controls, respectively =y(t) -y*(t) is the tracking or regulation (i.e. y* = 0) error = y(t) -y(t) is the prediction error = y*(t) -y(t) is the error due to the input non-linearity adaptation error in the parameter adaptive algorithm regression vector and regression vector deleting the component u(t) vectors of true and updated parameters parametrizing the corresponding control laws; O'(t) is intermediate steps in the adaptation of O(t) when projections are used subvectors of 0 and O(t) where the parameter which affects u(t)has been deleted. sets of integers, positive integers and positive integers including zero, respectively maximum and minimum eigenvalues of the ( . )-matrix, respectively