Abstract— This work concerns the development of a feedforward control strategy for measurable disturbance rejection in general nonlinear systems with feedback control. In other words, a separate design of disturbance feedforward controller is carried out to be augmented to an existing nonlinear feedback control system. With reference to a well known disturbance feedforward control of linear systems, feedforward control problem for general nonlinear control systems is formulated. The notion of disturbance effect is introduced with the aid of Lyapunov stability analysis of an asymptotically stable disturbance free nonlinear system with feedback control. With the establishment of the disturbance effect description, a feedforward control scheme that guarantees asymptotic stability in the presence of external disturbance is proposed and verified using the inverted pendulum system as an illustrative example. I. INTRODUCTION N a feedback control system with external disturbance, corrective action to eliminate the disturbance effects is only taken after the process has been affected. If the disturbance is measurable, feedforward control can be an effective open-loop control technique to prevent undesirable response due to the ability to take corrective actions before the disturbance affects the system. Feedforward control has been widely used in industry [1] and this efficient approach for disturbance rejection has motivated extensive research in the combination of feedback and feedforward control methods [2]-[7]. Some significant results in feedforward control involving nonlinear systems will be reviewed here. In [8], discrete- time feedforward/feedback controllers are developed for general nonlinear processes with stable zero dynamics and its connections with model predictive approaches are established. It is claimed that the derived controllers could reduce if not eliminate completely the effect of measurable disturbances and produce a prespecified linear response with respect to a reference input. The design of the controllers is synthesized in a coupled manner where separate objectives of the feedforward and feedback controllers are realized by means of one unified control law. In a recent work, an adaptive neural network feedforward compensator for external disturbances affecting a specific class of closed- Manuscript received April 15, 2009. This work was supported by the Ministry of Higher Education Malaysia and Universiti Teknologi Malaysia. All the authors are with Section for Automation & Control, Department of Electronic Systems, Aalborg University, Aalborg, Denmark (e-mail: {kad, alc, bisgaard}@es.aau.dk). loop system is introduced [9]. A nonlinear disturbance model is estimated to counteract the disturbance affecting a linear discrete closed-loop system. In this scheme, the disturbance model output is required to match the compensation signal for an effective rejection of the disturbance. Another feedforward control scheme using artificial neural networks is reported in [10] describing a nonlinear adaptive feedforward controller for compensation of external load disturbances in the idle speed control of an automotive engine. The feedforward only approach is based on Radial Basis Function network approximation of certain input-output mappings describing the system. The mapping involves optimal control input and a system variable that minimizes a quadratic performance index as the control objective. In another related work, the problem of adaptive feedforward compensation for input-to-state (and locally exponentially) convergent nonlinear systems is investigated [11]. The proposed scheme achieved disturbance rejection of a harmonic disturbance at the input of the nonlinear system class. In this paper, preliminary results concerning feedforward control for external disturbance rejection in general nonlinear feedback control systems is presented. Given an asymptotically controllable disturbance free nonlinear system with feedback control, a decoupled design of feedforward control is accomplished for disturbance elimination and hence to guarantee asymptotic controllability in the face of disturbance. Before a strategy could be developed involving measuring the disturbance and generating appropriate feedforward control signals, the effects of the disturbance on the system has to be well understood. For better understanding of the concept and to be aware of the objective of a feedforward control, the notion of disturbance effect in a nonlinear system affected by an external disturbance is introduced using Lyapunov stability analysis. A feedforward control strategy therefore can be one that nullifies the disturbance effect completely. It is shown that the proposed feedforward controller (to assist the feedback controller) that measures current disturbance and generates a corrective control signal warrants a decreasing control-Lyapunov function to achieve asymptotic controllability. The paper is organized as follows: Section II introduces preliminary background on nonlinear dynamical systems and classical approach of disturbance feedforward control for linear systems with additive disturbance, after which the Disturbance Effects in Nonlinear Control Systems and Feedforward Control Strategy Kumeresan A. Danapalasingam, Anders la Cour-Harbo and Morten Bisgaard I 2009 IEEE International Conference on Control and Automation Christchurch, New Zealand, December 9-11, 2009 FrMPo2.8 978-1-4244-4707-7/09/$25.00 ©2009 IEEE 1974