A control theory for Cartesian flexible robot arms (Q1270934): Difference between revisions

This paper makes a contribution to the mathematical modelling and control of flexible link robotic manipulators. The mathematics involved here relies on the theory of infinite dimensional control systems in Hilbert spaces. A Cartesian manipulator is considered, equiped with a long, flexible arm. Deflections of this arm are allowed in two orthogonal directions. A model of the arm's dynamics includes fourth order partial differential equations with boundary conditions, taking the following form \[ {d^2\over dt^2}y(t) +Ay(t)= bu(t),\tag \# \] where \(y(t)\) is an element of a Hilbert space, \(u(t)\) denotes a control and \(A\) is a linear operator. Several spectral properties of \(A\) have been studied. Then the model \((\#)\) has been converted into the form \[ {d\over dt}z(t)= Az(t)+ B(t)\tag \#\# \] with operator \(A\) defined again on a Hilbert space. For the last system, existence and uniqueness of solutions are shown, and approximate controllability is established. Endowed with a suitable output function, the system \((\#\#)\) is also proved to be observable. Eventually, static and dynamic feedback control laws are proposed, stabilising the end-effector position of the robot. In both cases exponential stability has been proved. The contents of this paper partially overlaps with the following recently published paper: \textit{W. Krabs} and \textit{N. Chi-Long}, Math. Methods Appl. Sci. 21, No. 1, 25-42 (1998; Zbl 0903.93002).