SYMBOLIC COMPUTATION IN NONLINEAR CONTROL-SYSTEM ANALYSIS AND DESIGN

Symbolic computation, also known as computer algebra, is a powerful to ol in solving rough and intricate problems in applied mathematics. We present a closely lied set of symbolic computation functions, together called the NON(sic)CON package, that solves some problems in the anal ysis and design of nonlinear control system. A model of the system mus t be available for the computation. The model, in general a set of non linear differential and algebraic equations in the state, input, and o utput of the system, is not required to be affine in the input nor io have a well-defined relative degree. The functions available range fro m the computation of the zero dynamics of the model to computing invar iant manifolds. All functions are based on constructive algorithms and are implemented in MAPLE. The NON(sic)CON package is successful for l ow order models that do not contain large intricate expressions. For l arge order models the package is less successful, due to an intrinsic limitation in MAPLE's handling of objects. Other problems stem from th e need to solve sets of nonlinear ((partial) differential) equations. The examples presented use interesting models of mechanical systems, i .e., a string of mass-damper-spring sets and a four-wheel vehicle.