Controlling a flexible robot arm driven by McKibben artificial muscles with direct transmission is delicate. The usual PID controller rapidly reveals itself to be inadequate and robust control tools are unavoidable. Classical sliding control, although robust, generates chatter. Several solutions are available to attenuate this phenomenon, among them the twisting and super twisting algorithms, which belong to the 2-sliding control set. It will be shown when to use the equivalent control and the effect of a noised sensor signal on control performance. Also, the use of an additional discontinuous term that increases robustness, performance and stability is put forward. Experimental results are presented and discussed.

Driving an anthropomorphic robot arm driven by pneumatic artificial rubber muscles (PARMs) with a direct transmission is quite delicate. PARMs present complex nonlin-earities and robust control tools are unevitable. Among them, those applied are the twisting and super twisting algorithms, which belong to the 2-higher order sliding mode control set. In this paper it is studied the effect of the equivalent control in sliding mode controller based on variable structure systems (VSS) theory. It will be shown when to use the equivalent control, and the effects of a noisy sensor signal on control performance. Experimental results are presented and discussed.

Pneumatic artificial rubber muscles (Parms) are similar to biological muscles as both act as springs. Stiffness varies with pressure for the Parm and with neural impulse for the biological muscle. As a Parm system is highly non linear, its classical control does not reveal to be totally indicated. Robust control techniques such as variable structure controls generating sliding modes are quite indicated. However, sliding mode control although robust generates chattering as the control switches across the sliding manifold. In this work, two controls are applied onto a robot driven by Parms in order to reduce this undesirable chatter. The first law is a generalised variable structure algorithm while the second one is a 2-sliding law, the “twisting” algorithm. Experimental results are presented and discussed.

McKibben artificial muscle is one of the most interesting artificial muscle which can actually serves as a biomimetic actuator for human-like robot limbs. However its peculiar non-linearities as the need to use it in direct drive to preserve its human-like compliance make very difficult the control of robots actuated by McKibben muscles. Experimental tests with a second-order sliding mode applied to the control of the regional structure of a 7R-anthropomorphic arm are reported. The so-called twisting algorithm appears particularly relevant for the combined positioning of joints submitted to gravity without inducing a chattering phenomenon. The role of an additional equivalent control term is discussed and it is shown to be efficient for facilitating motion of primary regional joints. In reported experiments with human-size arm and forearm in shoulder and elbow flexion, a response time without and with load between 2 and 3 s is obtained with a steady-state error lower than 0.5 degree.

We are concerned with the control of a 3-DOF robot arm actuated by pneumatic rubber muscles. The system is highly non-linear and somehow difficult to model therefore resorting to robust control is required. The work in this paper addresses this problem by presenting two types of robust control. One uses neural network control, which has powerful learning capability, adaptation and tackles nonlinearities; in our work the learning performed on-line is based on a binary reinforcement signal without knowing the nonlinearities appearing in the system and no preliminary off-line learning phase is required. The other control law is a Classical variable structure which is robust against parameters variations and external disturbances. Experimental results together with a comparative study are presented and discussed.

We are concerned with the control of a 3-DOF robot actuated by pneumatic rubber muscles. The system is highly non-linear and somehow difficult to model; variable structure system control imposes itself as a very relevant type of control. This paper presents two types of robust controls. One uses a classical variable structure (CVS) control associated to a PID control; the other is a derivative control. Both controllers are implemented on a robot driven by artificial rubber muscles. The goal is the comparison of the two controllers in their ability to reduce chatter. Experimental results are presented and discussed.

Pneumatic artificial rubber muscles (Parms) show similarity to biological muscles. A 2-sliding control technique is applied to the first two degrees of freedom of a robot actuated by such muscles. The Parms are arranged in opposite pair or antagonistic configuration. The objective is to show that without the use of the equivalent control, it is still possible to control the robot and at the same time reduce the chatter. Experimental results are presented.

It is known that classical sliding mode control generates chatter which is undesirable. One way to reduce this chatter is the use of high-order sliding mode (HOSM) control. The HOSM control techniques are applied to the first 2 d.o.f. of a robot actuated by pneumatic artificial rubber muscles (PARMs). The PARMs are arranged in opposite pairs (antagonistic configuration). The objective is to show that without the use of the equivalent control, it is still possible to control the robot by the use of HOSM and at the same time reduce the chatter. Experimental results are presented and comparison between two second-order sliding controls established.

This paper shows a robust control of a robot arm actuated by pneumatic artificial antagonistic muscles. The goal is chatter reduction.In this work the control problem is considered using a generalised variable structure algorithm which is similar to a 2-sliding mode control applied on a system of relative degree one with respect to the sliding constraint.The paper purpose is an experimental study where the control performances are discussed.