Suttipan Limanond

Neural network-based control design: an LMI approach

In this paper, we address a neural-network-based control design for a discrete-time nonlinear system. Our design approach is to approximate the nonlinear system with a multilayer perceptron of which the activation functions are of the sigmoid type symmetric to the origin. A linear difference inclusion representation is then established for this class of approximating neural networks and is used to design a state-feedback control law for the nonlinear system based on the certainty equivalence principle. The control design equations are shown to be a set of linear matrix inequalities where a convex optimization algorithm can be applied to determine the control signal. Further, the stability of the closed-loop is guaranteed in the sense that there exists a unique global attraction region in the neighborhood of the origin to which every trajectory of the closed-loop system converges. Finally, a simple example is presented so as to illustrate our control design procedure.

Monitoring and control of semiconductor manufacturing processes

Concerns optical measurement techniques for semiconductor manufacturing process monitoring and control. They can provide previously impossible real-time monitoring of several process variables, in-situ or ex-situ. This further enables the applications of more sophisticated real-time control algorithms, other than SPC. The SPC-based run-to-run (RtR) control, on the other hand, is still an instrumental part of the control algorithm, whenever there is a lack of real-time sensors for measuring critical process metrics. An emerging practice is to integrate RtR with the real-time control design to provide a comprehensive control design algorithm for semiconductor manufacture. The continued trend of semiconductor industry is toward an bigger wafers and smaller devices. This requires more integrated supervisory control, able to provide even tighter control, especially for photolithography, CVD, and etch processes. A unified framework needs to be established, at least for these critical processes, to facilitate and expedite systematic control design and development. In addition, as more sophisticated algorithms are being implemented, the control-related software and hardware should be user-friendly so that it can be operated by nonexpert personnel. Finally, since chip manufacturing consists of various processes, a comprehensive control algorithm on a factory-wide basis should utilize information (process-state, wafer-state, and tool-state data) from the current process as well as upstream and downstream processes.

Neural optimal etch time controller for reactive ion etching

In this article we address a neural network-based end point detection scheme for reactive ion etching process. We use the etch time as the critical parameter to indicate process end point. Further, our approach involves the use of a neural network-based predictive model relating various in situ measurements and end point detection signal to the resulting film thickness remaining in combination with an optimization algorithm. This circumvents the need for monitoring and operating on noisy end point detection signal typically associated with conventional detection schemes. Finally, we present simulation studies based on production data to further demonstrate the associated design procedures and the feasibility of the algorithm.

Model reference adaptive and non-adaptive control of multivariable linear time-varying plants: the exact matching case

In this paper we address the adaptive and nonadaptive model reference control problem for a class of linear time-varying plants, namely the index-invariant ones. We show that, under appropriate controllability and observability conditions, this class of plants admits a fractional description in terms of polynomial differential operators and, as such, allows for a polynomial equation-based controller design. We also show that, for a control objective of the model reference type, the controller can be designed by solving a set of algebraic equations. Further, in the case where the plant parameters are only partially known, we employ a gradient-based adaptive law with projection and normalization to update the controller parameters and establish the stability and tracking properties of adaptive closed-loop plant.<<ETX>>

Adaptive control of time-varying systems: an application to the attitude-momentum control of the space station

The authors consider an application of adaptive techniques to control the space station in orbit with a moving payload. The nonadaptive counterpart of this problem has been considered by B. Wie et al. (1984). Such a controller, designed to tolerate constant parametric uncertainty, may exhibit unsatisfactory behavior when the variation of the system parameters is not sufficiently slow. Motivated by this practical problem, the theory and application of adaptive techniques in up-dating the parameters of the controller designed by Wie et al. is discussed. The main theoretical problem arising in this application is associated with the adaptive identification and control of multivariable linear time-varying systems. Other theoretical problems arise from the decentralized nature of the controller of Wie et al., and a hybrid implementation of the adaptive control scheme to reduce the required computational effort. Such problems have been studied individually in the literature, and are integrated in the present study to provide a complete theoretical justification of the adaptive approach to the application at hand.<<ETX>>

RIE process time optimization from real time sensory data

Abstract This paper addresses the issue of end point detection and etch time control for a reactive ion etch process. Our approach involves the use of neural networks to model functional relationship between an end point detection signal as well as various in situ measurements, and the resulting film thickness remaining. An optimization algorithm is then employed to determine the optimal etch time based on the neural network model in order to achieve the desired level of film thickness remaining. This circumvents the need for monitoring and operating on noisy end point detection signal typically associated with conventional detection schemes. Furthermore, a self-organizing map is incorporated as a process classifier for identifying the current operating condition of the reactive ion etch process, which provides an alternative for tracking process variations and uncertainties, without the need tor the on-line adaptation of model or controller parameters. Tested on data from 77 randomly selected wafers, our controller yields a film thickness distribution with the standard deviation of 6.54, a 50% improvement over the scheme currently implemented in production.