Abstracts 132 A Design Algoritiun for Anti-Windup Compensators -
Polynomial Approach S. Riinnbtiek, J. Sternby, pp 619-624 This paper discusses how an LTI "polynomial controller" for a SISO-system with actuator constraints can be designed using a conventional linear pole-placement technique. This becomes possible due to the utilization of an anti-windup technique, in which the effect of control-signal saturation is interpreted as being caused by an equivalent disturbance. The results are summarized into a design algorithm that supports the introduction of arbitrary specific factors into the controller polynomials. The technique can be used for both continuons-time and discrete-time controllers. 133 Experimental Verification of Multivarlable Transfer Functions for a Slewing Piezoelectric Lamhmte Beam T.E. Alberts, T.V. DuBois, H.R. Pota, pp 625-628 This paper presents experimental verification of a distributedparameter model for a slewing beam system with piezoelectric actuators end sensors. The beam is pinned at the proximal end, an endpoint motion sensor is a_n_achedat the distal end, and patches of thin piezoelectric laminates are attached to its surface. Experimental results, which verify the model, are presented using a beam experiment at the US Air Force Academy, Frank J. Seiler Research laboratory.
134 Instrumentation, Control and Management of Batch Reactors Using Distributed Controls A. Kaya, A. Kumar, J. Glass, pp 629-632 The work reported here demons~ates the implementation of various instrumentation, control, and management functions by using a distributed control system for a batch reactor. Starting with smart instruments (transmitters), feedforward, cascade, adaptive features of feedback control are configured. Facilities for user convenience and management ease are included, in addition to safety, alarm and control integrity. The system uses a reactor simulator to predict the results of reaction in new recipes. Complete data-acquisition, -logging, and -storage capabilities are resident in the control room and remotely accessed by PCs. This work demonstrates the capabilities of state-of-the-art distributed controls, used in a pilot reactor plant. 135 Dynamic Generation of a Simplified Model of the Environment for AGVs Control J. Pages, A. Casals, J. Amat, pp 633-636 The use of antomatic guided vehicles in industry is increasing with the advances and growth of flexible manufacturing systems. There exists a trade off between their expected degree of autonomy, and their complexity and cost. This paper proposes a low-cost solution to AGV guidance by means of a simplified model of the mobile environment. Some constraints that fit with real industrial environments are imposed, to attain a good compromise between cost and system speed. 136 Variable Structure Pl Controllers in the Presence of Uncertaimy and Saturation C. ScaB, G. Nardi, A. Landi, A. Balestrino, pp 63%642 Performance of variable-structure controllers, with a PI algorithm varying as a function of the error, is analyzed in the case of systems subject to uncertainty and satmation nonlinearities. The comparison is carried out with respect to constant-parameter controllers, designed according to the internal model control theory and supplied with an anti-windnp device to limit the deterioration of performance caused by saturation. Simulation results show that for comparable speed of the time responses in the nominal and unconstrained case, the VS-PI controller is less affected by uncertainty and saturation, both for set-point tracking and for the rejection of disturbances entering at the plant input. 137 Qualitative Simulation of an Inflation Model Y.K. Wong, R.R. Leitch, G.J. Wyatt, p p 643-646 For many tasks it may be more appropriate to reason with qualitative models which are an inexact representation of the real world, than to use the more-powerful classical tools of differential or difference equations. Qualitative methods allow one to build models which do not necessarily incorporate assumptions like linearity or constancy over time. Even with such imprecise
131
knowledge, there is enough information in a qualitative description to support qualitative simulation, which predicts the possible qualitative behaviour of economic systems. 138 Interval Input-Output Model
M.S. Koparanova, S.M. Koparanov, pp 647-650 The difficulties in constructing detailed input-output tables bring about the use of interval data as more reliable for input-output analysis. It is for this reason that an interval interpretation of the static input-output model is introduced, and an interval estimation of Leontiefs matrix is proposed. The extension presented in the paper is cbarncterised by non-linear relations of inter-industry economics, interval interpretation of the a priori information, and interval endogenous parameters. The main assumptions and a necessary and sufficient condition for the existence of a solution of the interval input-output model (IIOM) are constructed. A method for estimating input-output coefficients using interval data is proposed.
139 Output and Price Determination: A Theoretical and Empirical Study A. Fusari, p p 651-654 This paper develops a model aimed at representing the functioning of the whole economic mechanism with a high degree of detail, and avoiding some typically unrealistic assumptions of general equilibrium models, both of the Walras and Arrow-Debreu or yon Neuman kind. Some econometric estimation allows a better understanding of the reliability and usefulness for applied analysis of the theory proposed.
140 Fuzzy Sets in the Development of Policies for International Events C.D. Antonini, pp 655-658 Events in international relations are assumed to be related, and their relationship is supposed to be structtwed. The description of this structure is usually given in linguistic terms, although there may also be numerical values that may be used to re~eseut it. In this work, fuzzy sets are used to symbolize the assignment of relationships among the elements of the structure, and fuzzy logic to manipulate the linguistic variables and numerical values that enter in the description of the structure. Difficulties related to any approach directed at formalizing this relationship will be indicated, as well as areas of further research.
141 Treod Towards Open Architecture Control Systems and Its lmpileations for Developing Countries C.G. Park, p p 659-663 Process and manufacturing automation has been a direct beneficiary of recent advances in computer technology. However, the heavy demands of process and manufacturing industries for refiability and long service life have allowed only a few vendors to finance the development of new systems requiring long development cycles and great amounts of resources. Once new systems are developed, it is in the vendors' interest to protect their investment by keeping them proprietary. Recent trends in the computer industry toward open systems, however, are changing the way new technologies are adopted by the process and manufacturing industries. This provides opportunities for developing countries to accelerate automation by embracing open systems 142 Decision Support Tools Using Hierarchical Multicriteria Models B. GaiHy, F. Bonller, M~ lnstaBe, pp 665-667 The purpose of this paper is to study a special class of problems where systems are not driven by classical transfer functions or state equations, but by an impficit relation between the state, the input and the output of the system. This relation is based upon a certain f~eedom left to subsystems to react according to their own objectives. An original theoretical approach is presented, and a hierarchical multicriterion and multiperson model of decision making is proposed, where the motivations of the people involved are dealt with through the use of incitements and/or incentive strategies.