Application of Microcomputer-Based Model Reference Adaptive Control to a Hydraulic Positioning System

h unbehauen,peijun du,ulrich keuchel

DOI: https://doi.org/10.1016/S1474-6670(17)52740-8

1990-01-01

IFAC Proceedings Volumes

Abstract:Publisher Summary This chapter discusses the application of a microcomputer-based model reference adaptive control to a hydraulic positioning system model. A direct adaptive control scheme based on the model reference principle guarantees stable operation. This is achieved by augmenting the plant output signal using a so called correction network connected in parallel to the plant, which acts as a dynamic weighting filter for the manipulating signal. This network is designed using root locus or pole assignment theory. For parameter estimation, several methods and algorithms such as stochastic approximation and recursive least squares have been employed. Only numerically stable update algorithms such as square-root filtering or UDUT factorization of the covariance matrix are used. The chapter presents some experiments that show that adaptive model reference control assures zero steady-state error for non-negligible load, even in the absence of integral action in the control law.