Cecil L. Smith

Digital control of industrial processes

The application of computers for controlling process units is now about tell years old. The applications have progressed from the experimental stages of early development to a well-established (economically speaking) control practice. This has not been without some anxious moments and a few outright failures. These have not been forgotten, but current successes are proving beyond any doubt that digital conputers are capable of generating an attractive economic return on the investment required for their installation. This paper has three objectives: first, to acquaint those outside the process industries of the general process control problem in order that they may appreciate some of the requirements imposed on a control computer; second, to describe the computer hardware normally considered for process control; third, to present the general philosophy of the real-time software systems required for process control.

An optimal controller for multivariable systems subject to disturbance inputs

Abstract : Most industrial processes have two features in common: (1) they are multivariate in nature and (2) they are subject to unmeasured disturbances for which control systems must be installed. The conventional design of multivariable systems has been considered by Bollinger and Lamb, Dahlin et. al and Shinskey. Douglas considered the constrained optimal control of multivariable systems, deriving the optimal switching curves for a representative system. This objective of this work is to present a design technique for linear, continuous controllers for multivariable process systems. The method will use the concepts for the optimal control of linear systems as presented by Athans and Falb. The conventional disturbance control problem is reformulated in a manner that allows optimal control theory to be used directly. (Author)

On-Line Model Identification and Control Using the Kalman Filter.

Abstract : For some systems a model determined off-line can be satisfactorily used to design control systems. However, for many systems, the best values for the model parameters will change as the system operation varies. For effective control these parameter changes must be identified on-line and incorporated into the digital control strategy. Equally as important as the identification of changing parameters is the identification of unmeasured disturbances which upset the system operation. Ideally, the identification of model parameters and unmeasured disturbances would be accomplished by measuring only the controlled variable. One method which holds promise is the extended Kalman filter. The Kalman filter, which was developed for rejecting noise from measurements, is applied in this paper to problems of parameters and disturbance identification and state estimation of first- and second-order processes. The primary objective is to evaluate its performance for estimating the unmeasured disturbances encountered so frequently in operating control systems. (Author)

Dynamic simulation in the design and analysis of water reuse units

Abstract To provide the flexibility to meet a range of requirements in the design and analysis of water resuse units, a general purpose dynamic process simulator tailored to water reuse application has been developed. Because many wastewater streams are characterized by time-varying flows and concentrations, unsteady state considerations are and essential part of the analysis. Model parameters for the simulation are based on data obtained from experimental tests on the performance of the various unit processes on the wastewater to be processed. These model parameters and the plant configuration comprise the input data to the simulator.

A dynamic model of a water treatment unit

For support of field medical units, the U.S. army is developing a self-contained waste water treatment system to produce portable water for use within the unit. To aid in developing the control strategy and the fault detection/fault isolation logic, a dynamic model of the system is being developed. The first phase of the effort is to develop models of the systems components, such a ultrafiltration, reverse osmosis, ozonation, etc. These models are then combined to obtain a model for the complete system. To the greatest extent possible, fundamental equations form the basis of the models. In determining coefficients for the models, the the data was obtained from the experiments that formed the basis of the design of a pilot plant version of the unit. To refine the model, this data is then supplemented by operational data from the pilot plant.

Criteria for evaluating Fourier transform computational results

Abstract : When numerical methods are used to calculate Fourier transforms from discrete points from the input and output of a dynamic system, the resulting Bode plot degenerates at high frequencies. Several factors contribute to this, e.g., foldover, noise in data, computational methods, etc. To effectively analyze the effect of these, it is essential that the blow-up point be determined accurately and consistently. This paper examines integral criteria for this purpose. (Author)

Calculation of temperature profile in ablator char using MIMIC.

The differential equation describing the temperature pro file in the carbon zone of a charring ablator was simulated on an IBM 7040 digital computer using MIMIC, a digital- analog simulator language. This differential equation de scribes a nonlinear, two-point boundary value problem, thus requiring an iterative procedure to converge to a solu tion. In order to handle the iterative procedure, the logical- control-variable feature of MIMIC was especially useful.