Predictive Control

Daniel Lequesne

• Publisher: ISTE Press - Elsevier
• Year: 2017
• Print ISBN: 9781785482625
• eBook ISBN: 9780081023884
• Topic: Technology & Engineering
• Subtopic: Engineering General
• Index: Technology & Engineering

Predictive Control is aimed at students wishing to learn predictive control, as well as teachers, engineers and technicians of the profession. The book proposes a simple predictive controller where the control laws are given in clear text that requires no calculations. Adjustment, reduced to one or two parameters, is particularly easy, by means of charts, thus allowing the operator to choose the horizon according to the desired performances. Implementation is discussed in detail in two forms: RS or RST controller in z-1, and pseudo-code realization algorithms for a complete program (model and controller).The book is simple and practical, with the aim of the industrial implementation of many processes: Broïda models, Strejc, integrators, dual integrators, with delay, or with inverse response. All settings are abundantly illustrated with response curves.- Present a practical guide to predictive control- Offers a simple predictive controller for a wide range of industrial applications- Summarizes, in tables, all the calculations that have been carried out to allow immediate implementation

1

Principle

Abstract

Predictive control is a general technique that relies essentially on the capacity to predict how processes behave in order to better correct them.

Keywords

Models; Prediction; Ramp prediction; Simulation; Stable systems; Trajectory

Predictive control is a general technique that relies essentially on the capacity to predict how processes behave in order to better correct them.

The correction is provided by a controller: this is the case for a classical control loop, in which the controller receives a setpoint value C and measure M of the quantity to be controlled; it provides a command U that controls the process. The loop is represented in Figure 1.1 where the input X symbolizes the eventual external disturbances.

Figure 1.1 Control loop

The specificity of predictive control resides in the calculation of what will be measured in the future and its consideration in order to determine the command for obtaining the given desired trajectory corresponding to a setpoint value. This results in the following points:

– it is not possible to know a future measurement unless a representation model of the process is available;

– a future moment implies a present moment: the controller is numerical, with a sampling period Te; at each Te, the command U is recalculated;

– the trajectory is defined by the user and it is a way of representing the performances to be obtained.

1.1 Trajectory

The trajectory is defined relative to the (desired) setpoint and the measure on the process as shown in Figure 1.2. Instant k is the present instant; the trajectory represents the desired evolution of the measure starting from its current value M(k) to a final value that is equal to the setpoint. The represented setpoint corresponds to the current instant, C(k), and it is supposed c...

Table of contents

• Cover
• Title page
• Table of Contents
• Copyright
• Preface
• 1: Principle
• 2: Control Law
• 3: Process Models
• 4: Implementation
• 5: Setting of Stable Systems
• 6: Setting of Integrating Systems
• 7: Performances and Setting
• Conclusion
• Appendix
• Bibliography
• Index