Adaptive control is now finding its way into the marketplace after many years of effort. This paper reviews some ideas used to design adaptive control systems. It covers early ideas which primarily attempt to compensate for gain variations and more general methods like gain scheduling, model reference adaptive control, and self-tuning regulators. It is shown that adaptive control laws can be obtained using stochastic control theory. Techniques for analyzing adaptive systems are discussed. This covers stability and convergence analysis. Issues of importance for applications like parameterization, tuning, and tracking, as well as different ways of using adaptive control are also discussed. An overview of applications which includes feasibility studies as well as products based on adaptive techniques concludes the paper.

/pdf/adaptive-feedback-control-4wzqer5bdp.pdf

Adaptive feedback control

Some of the ideas associated with process control systems have already been discussed in §1.4. The majority of processes may be controlled automatically by adjusting the rate of flow of energy or material entering the process. In manually-controlled systems the human operator adjusts the rate of flow by means of a hand-operated valve; he will receive information regarding the state of the process from a measuring instrument giving a visual indication. Manual control has numerous disadvantages and indeed manual control of many modern plants would be virtually impossible, owing to their complexity and the speed and precision of control they require for successful operation.

Process Control Systems

On-line optimal control of a seeded batch cooling crystallizer

Theory and Application of Adaptive Control

The authors have developed a reliable long-range predictor, comprising two neural networks with external feedback in series, and investigated its applicability for model predictive control on a simulation example. The networks use external feedback of the process state, yielding a state-space mapping that eliminates the drawbacks of the input-output mapping of the feedforward networks. This paper applies the long-range predictor to the model predictive control of an experimental bench-scale semi-batch chemical reactor. Examples of yield maximization for a reaction with complex kinetics are used to assess the proposed predictive control scheme. Control performance is compared for predictors based on the proposed external-feedback networks and on conventional feedforward networks. Results for various operating conditions, disturbances, and included analytical models demonstrate the superiority of the proposed control scheme in experiments.

Predictive control of a bench-scale chemical reactor based on neural-network models

Adaptive inferential control for chemical processes

A unified approach to discrete-time explicit model reference adaptive control schemes has been applied to the multivariable control of a pilot scale packed distillation column separating a mixture of methanol and water. The control scheme is based on Landau-Lozano's approach (1981), and a method of control costing of Clarke and Grawthrop (1975) and an estimation method of nonzero d.c. value are also incorporated. The effects of the model order and time delays are illustrated. The resulting control performance is compared with that achieved with conventional proportional plus integral action.

Experimental Investigation of Adaptive Control of a Distillation Column

An optimal control theory for a class of non-linear distributed parameter control systems in the general setting involving more than one spatial co-ordinate is developed for use on control of a typical soaking pit in a steel industry. The system is described by a set of non-linear partial differential equations in multidimensional spatial co-ordinates and non-linear boundary conditions with time-dependent boundary controls as well as a spatially independent parameter vector which is governed by its own set of dynamic equations. The set of necessary conditions for optimality obtained from the theoretical development is directly applied to the optimal heating control of the soaking pit with rectangular ingots. The numerical solution by iterations demonstrates the success of the technique and the algorithm leading to the optimal policy for heating control of a typical soaking pit with minimum fuel consumption.

Optimal control of temperature and fuel consumption for soaking pits in a steel plant—a distributed parameter approach

This paper describes a neural network approach to the automatic startup and control of a batch reactor where exothermic reactions take place. A backpropagation neural network model is used for online determination of the startup switching time and a model predictive control strategy based on a back propagation neural network is to be designed for a regulatory control after the desired reactor temperature is reached during startup. Simulation results are presented to illustrate the applicability of the proposed neural network approach to the rapid startup and control of exothermic batch reactors.

Neural Networks Approach to Automatic Startup and Control of An Exothermic Batch Reactor

Un effort est fait pour unifier 2 methodes d'obtention de predicteur analytique du point de vue de l'estimation des parametres mais aucune hypothese est emise sur la fonction de transfert des perturbations

Won-Kyoo Lee

Papers

Adaptive inferential control for chemical processes

Experimental Investigation of Adaptive Control of a Distillation Column

Optimal control of temperature and fuel consumption for soaking pits in a steel plant—a distributed parameter approach

Neural Networks Approach to Automatic Startup and Control of An Exothermic Batch Reactor

Generalized analytical predictor