INTELLIGENT HIERARCHICAL CONTROL SYSTEM

FOR COMPLEX PROCESSES

Three Levels Control System

Yuri V. Mitrishkin

Bauman Moscow State Technical University, Second Baumanskaya St.,5, Moscow, Russia

Rodolfo Haber Guerra

Instituto de Automática Industrial, CSIC, Madrid, Spain

Keywords: Complex dynamic systems, Feedback, Hierarchical, Robust, Adaptive, Self-organizing, Decision making,

Intelligent control, Plasma in tokamaks.

Abstract: The paper presents a concept of intelligent hierarchical control for complex dynamical processes and

suggests architecture of control system consisting of robust, adaptive, and self-organizing levels. Intelligent

features of the proposed system are mostly concentrated at self-organizing level incorporated into self-

learning, self-configuring, self-optimizing, and decision making algorithms. State-of-the-art at each level is

described. Case studies have been chosen from the area of plasma control in tokamak-reactors.

1 INTRODUCTION

Recent advances in control strategies,

communications, hard and soft-computing

technologies have favoured an increasing trend

towards the new generation of networked control

systems for complex processes. The proposal

described herein will address the development of

scalable control methods and systems in accordance

with the Information and Communication

Technologies (ICT) Work Programme, ICT-2009

3.5a: Foundations of complex systems engineering:

To achieve robust, predictable and self-adaptive

behavior for large-scale networked systems

characterized by complex dynamic behavior through

the development of novel abstractions and scalable

methods for sensing, control and decision-making.

The scope covers foundational multi-disciplinary

research and proof of concept addressing the whole

chain from modeling, sensing, monitoring and

actuation, to adaptive and cooperative control and

decision making (European Commission, 2008).

To meet the goal stated by the EC a three level

intelligent hierarchical control system was suggested

by Bauman Moscow State Technical University to

be applied to solve control problems of complex

dynamic processes in science, engineering, and

industry.

The project is focused on design and

development of scalar (Single-Input/Single-Output:

SISO) and multivariable (Multi-Input/Multi-Output

MIMO) networked control systems based on

scalable control algorithms for uncertain time-

varying nonlinear complex dynamic processes.

The major innovation of the proposal implies the

elaboration of a new methodology for designing

hierarchical adaptive self-organizing control systems

to be applied to complex production processes, such

as: plasma energy release, chemical and biological

processes, casting in metallurgy, oil refinery, and so

forth.

2 PHILOSOPHY OF

HIERARCHICAL CONTROL

Industry and academia have investigated a wide

range of decentralized control architectures ranging

from hierarchical decomposition to a completely

decentralized (heterarchical) approach where

individual controllers are assigned to subsystems

and may work independently or may share

333

V. Mitrishkin Y. and Haber Guerra R.

INTELLIGENT HIERARCHICAL CONTROL SYSTEM FOR COMPLEX PROCESSES - Three Levels Control System.

DOI: 10.5220/0002171003330336

In Proceedings of the 6th International Conference on Informatics in Control, Automation and Robotics (ICINCO 2009), page

ISBN: 978-989-8111-99-9

data/information.

The main disadvantage of heterarchical

approaches is that global optima cannot be

guaranteed and predictions of the system’s

behaviour can only be made at the aggregate level.

Hierarchical and heterarchical architectures lie at

opposite ends of the distributed control architectures

spectrum. The hierarchical approach is rigid and

suffers from many of the shortcomings of the

centralized approach, whereas it provides clear

advantages in terms of overall system coordination

alternatively. The heterarchical approach is flexible

and fault-tolerant, but arguably difficult to

coordinate.

Hierarchical control and supervision schemes

have been widely studied as a possible solution for

optimizing complex systems. A variety of schemes

can be implemented in order to profit from the

advantages of this architecture, and the applications

run all the way from fault-tolerant aircraft control

problems to servo systems supervision (Kwong et

al., 1995).

Hierarchical control allows any available data

from the low-level control system to be used at a

higher level to characterize the system’s current

behaviour. Moreover, hierarchical control can be

used to integrate extra information (in addition to

that concerning the usual control-loop variables such

as output, error, etc.) into the control decision-

making process. In many situations a hierarchical

approach is an advantageous option for process

optimization, instead of sophisticated design and

implementation of high-performance low-level

controllers, because the hierarchical approach can

compensate factors that are not taken into account in

the design of low-level controllers (Berenji et al.,

1991).

Thanks to its own structural essence, the

hierarchical control scheme ensures flexibility and

compatibility with other controllers that have

already been installed. It has other strong points as

well, such as the relatively low cost of investments

in improving automation scheme performance, the

possibility of exploiting already-installed low-level

regulation systems, and the relatively low cost of

measurement systems which makes hierarchical

control a wise choice from economic and practical

viewpoints.

The hierarchical methodology will cover three

basic levels, namely:

I) physical control level composed of controlled

process interacting with robust or classical controller

through sensors and actuators;

II) adaptive level that implements scalable

adaptation algorithms and enables the robust

controller to satisfy a number of time-varying

constraints;

III) knowledge-based self-organizing level

which executes a set of self-learning, self-

configuring, self-optimizing, and decision-making

algorithms allowing possible dynamic changes in the

system architecture aimed at optimal control and

dynamic reconfiguration of the robust controller and

making it easier to satisfy process-critical

constraints, such as respond to the deadlines,

saturations and so on.

Advanced algorithms designed for all control

levels I, II, III and their interactions provide a high

degree of system flexibility, robustness, accuracy,

reliability, dependability, and survivability to deal

with disruptive, uncertain, and unforeseen events in

automatic mode (without human intervention).

The philosophy of the three-levels intelligent

hierarchic control system proposed is a strategy of

the project to be organized. Complex processes

under control should be first of all thoroughly

investigated, then classical or robust control systems

are to be developed. After that, improvement of

lower level should be done by means of levels II and

III including decision making approach which may

be performed by experts at the beginning of system

design and not in automatic mode.

3 STATE-OF-THE-ART

The design of the main SISO and MIMO control

loops should be based on, in a certain sense,

classical approaches which have been used in

modeling and practice and demonstrated efficacy in

a number of applications. These approaches are

based on design principles for which reliable

dynamic processes models are created by means of

First Principle Equations methodology

(Khayrutdinov et al., 1993, Leonov et al., 2005),

identification (Mitrishkin, 2004), linearization and

subsequent reduction procedure. The resulting real

world models have been used in system closed-loop

to predict optimal control actions at each discrete

timing interval to achieve the control goal taking

into account input constraints (Mitrishkin and

Korostelev, 2008). Decoupling control leads to the

design of astatic multivariable systems (Leonov et

al., 2005). A number of optimization approaches,

specifically off-line Linear Quadratic techniques

(Belyakov et al., 1999),

∞

and μ synthesis

ICINCO 2009 - 6th International Conference on Informatics in Control, Automation and Robotics

334

(Mitrishkin et al., 2003, Ariola and Pironti, 2008) as

well as online automatic optimization of extremum

search (Mitrishkin, 2004), gave acceptable results.

Adaptive Kalman filter made it possible to estimate

process parameters in real time and gave a chance to

adapt the system by adaptive algorithm to time-

varying process parameters (Mitrishkin and

Kuznetsov, 1993). As this took place, an external

disturbance was estimated and estimation value was

used to compensate the disturbance itself

(Mitrishkin, 2004).

Some results in the topics in relation to this

proposal are the design and implementation of:

intelligent hierarchical control and supervisory

systems (Peres et al., 1999); control strategies by

fuzzy, neural, neuro-fuzzy, and evolutionary neuro-

fuzzy internal model control systems (Haber et al.,

2004); rapid control prototyping for networked and

embedded control (Haber et al., 2008); embedded

intelligent control systems in open architectures

(Haber and Alique, 2007), and networked control and

intelligent monitoring for macro- and micro-scale

manufacturing processes (Haber et al., 2008).

The synthesis of intelligent hierarchical control

system (Andrikov and Konykov, 2004) was applied

to improve car braking controllability by H

∞

control

theory.

4 STATEMENTS OF CONTROL

PROBLEMS

A number of new important complex control

problems have to be studied, discussed and

formulated to achieve control goals of acceptable

trade-off between robust stability and performance

of feedback systems. The problem statements

concern the stabilization and tracking process output

signals, optimal distribution of process parameters in

space in the presence of non-modeled process

dynamics, unobserved disturbances, nonlinearities,

in particular saturations, wideband insufficiently

known noise in output signals, non-minimum-phase

dynamics, and time-varying parameters. To solve

these control problems a set of approaches from

linear and nonlinear control theory will be explored

and developed to achieve scalable

∞

robust,

decoupling, model predictive, adaptive, hierarchy,

cascade, soft-computing based control (e.g., neuro-

fuzzy control systems), and facilitate decision-

making in new appropriate combinations within

continuous and discrete time of the three-level

hierarchic control system. Scalable control

algorithms mean that the algorithms may be

generalized to any numbers of controlled plant

inputs, outputs, and space states.

5 PRACTICAL APPLICATIONS

In order to validate the suggested approaches plasma

energy release case study is planned to be

investigated. Control methodologies will be applied

to plasma vertically unstable position, shape, and

current in the presence of voltages and current

saturations in poloidal magnetic coils in ITER

(International Thermonuclear Experimental Reactor,

www.iter.org). Multivariable robust controller

design (level I) with adaptation on the level II is

proposed to be done for the whole plasma discharge

of plasma current ramp-up, ramp-down, and at quasi

stationary stages. It is planned to be applied for

ITER reference scenario No 2 with plasma current

on flat-top of 15 MA and for reversed share scenario

No 4 of plasma current of 9 MA. Mathematical

modeling of control systems to be developed on

plasma-physics code DINA (Khayrutdinov et al.,

1993) is assumed to be fulfilled in tracking and

stabilizing modes at disturbances of minor

disruption type.

The project control methodologies are planned to

be applied to solve plasma kinetic control problem

as well. Plasma kinetic control means creation and

maintenance of optimal plasma current, temperature,

and density profiles by means of additional heating

sources. Such regimes are necessary for stationary

operation of tokamak-reactors. Development of

kinetic plasma models of plasma current,

temperature, and density profiles and their

identification are supposed to be created. Then

design and modeling of plasma profiles control

systems are assumed to be performed.

The final issue of this activity is integration of

plasma magnetic and kinetic control systems.

In the process of hierarchical structure control

systems design the synthesis, analysis, and

numerical modeling approaches are proposed to be

performed in MATLAB/SIMULINK environment.

The ITER functionality is planned to be

performed by means of CODAC software

specifically Control, Data Access, and

Communications (www.iter.org) via which one can

install hierarchical control scheme proposed.

INTELLIGENT HIERARCHICAL CONTROL SYSTEM FOR COMPLEX PROCESSES - Three Levels Control System

335

6 CONCLUSIONS

The concept of three levels hierarchic control system

was presented and discussed namely: architectural

details, state-of-the-art, statement of control

problems, and practical applications.

The project will result in the creation of new

process models, procedures of their identification

and reduction, efficient, robust, predictable, and safe

ICT control methodologies, scalable control

algorithms, and high-performance controllers with

reconfigurable architecture for the problem oriented

hierarchical systems under consideration. Scientific,

engineering, and industrial results will be

accumulated in the data and knowledge bases with

accurate classification, qualitative and quantitative

assessment, and generalization.

REFERENCES

Albus J. S., 1991. Outline for a theory of intelligence,

IEEE Trans. Syst., Man, Cybern. A Vol. 21 (3), pp.

473-508.

Andrikov D., Konykov V., 2004. Robust Н

∞

– optimal

controller for car with ABS in emergency situation in

slip mode. Herald of BMSU, Instrument engineering

series. Vol. 57, No. 4, pp. 44–57 (in Russian).

Ariola M., Pironty A., 2008. Magnetic Control of

Tokamak Plasmas. Springer-Verlag.

Belyakov V., Kavin A., Kharitonov V., Misenov B.,

Mitrishkin Y. et al. 1999. Linear Quadratic Gaussian

Controller Design for Plasma Current, Position and

Shape Control System in ITER, Fusion Engineering

and Design, Vol. 45, pp. 55-64.

Berenji H. R., Chen Y.-Y., Lee C.-C., Yang J.-S.,

Murugesan S., 1991. A hierarchical approach to

designing approximate reasoning-based controllers for

dynamic physical systems, in Uncertainty in Artificial

Intelligence Vol. 6, pp. 331-343, 1991.

European Commission C (2008)6827, 17 November 2008.

Work Programme 2009. Cooperation, Theme 3. ICT –

Information and Communication Technologies, p. 38.

Haber R. E., Alique J. R., 2004. Nonlinear internal model

control using neural networks: an application for

machining processes. Neural Computing &

Applications, vol. 13, pp. 47-55.

Haber R.E., Villena P., Haber-Haber R., Alique J.R.,

2008. Fast design and implementation of intelligent

controllers. DYNA, vol. 83 (8), pp. 127-134.

Haber R. E., Alique J. R., 2007. Fuzzy logic-based torque

control system for milling process optimization. IEEE

Trans. on Systems Man and Cybernetics. Part C-

Applications and Reviews, vol. 37, pp. 941-950.

Haber R. E., Martin D., Haber-Haber R., Alique A., 2008.

Networked fuzzy control system for a high-

performance drilling process. Journal of

Manufacturing Science and Engineering-Trans. of the

ASME, vol. 130, pp. 68-75.

Khayrutdinov R.R., Lukash V.E., 1993. Studies of Plasma

Equilibrium and Transport in a Tokamak Fusion

Device with the Inverse-Variable Technique. Journal

Comp. Physics, Vol. 109, pp. 193–201.

Kwong W. A., Passino K.M., Laukonen E.G., Yurkovich

S., 1995. Expert supervision of fuzzy learning systems

for fault tolerant aircraft control, Proc. of IEEE Vol.

83 (3), pp. 466-483.

Leonov V., Mitrishkin Y., Zhogolev V., 2005. Simulation

of Burning ITER Plasma in Multi-Variable Kinetic

Control System. Proc. of 32nd Plasma Physics Conf.

of European Physics Society, Tarragona, Spain, ID

P5.078.

Mitrishkin Y., Kuznetsov E., 1993. Estimation of

Parameters of Stabilized Plasma. Plasma Devices and

Operations, No. 3, Vol. 2, pp. 277-286.

Mitrishkin Y., Kurachi K., Kimura H., 2003. Plasma

multivariable robust control system design and

simulation for a thermonuclear tokamak-reactor,

International Journal of Control, Vol. 76, No. 13, pp.

1358-1374.

Mitrishkin, Y., 2004. Comprehensive Design and

Implementation of Plasma Adaptive Self-Oscillations

and Robust Control Systems in Thermonuclear

Installations. Proc. of 8

World Multi-Conference on

Systemics, Cybernetics and Informatics, Orlando, FL,

USA, Vol. XV, pp. 247-252.

Mitrishkin Y., Korostelev A., 2008. System with

Predictive Model for Plasma Shape and Current

Control in Tokamak. Control Sciences, No.5, pp. 22-

34 (in Russian).

Peres C. R., Haber R. E., Haber R. H., Alique A., Ros S.,

1999. Fuzzy model and hierarchical fuzzy control

integration: an approach for milling process

optimization. Computers in Industry, vol. 39, pp. 199-

207.

ICINCO 2009 - 6th International Conference on Informatics in Control, Automation and Robotics

336