Autonomous control of complex systems: robotic applications

https://doi.org/10.1016/S0096-3003(99)00285-4Get rights and content

Abstract

One of the biggest challenges of any control paradigm is being able to handle large complex systems under unforeseen uncertainties. A system may be called complex here if its dimension (order) is too high and its model (if available) is nonlinear, interconnected, and information on the system is uncertain such that classical techniques cannot easily handle the problem. Soft computing, a collection of fuzzy logic, neuro-computing, genetic algorithms and genetic programming, has proven to be a powerful tool for adding autonomy to many complex systems. For such systems the size soft computing control architecture will be nearly infinite. Examples of complex systems are power networks, national air traffic control system, an integrated manufacturing plant, etc. In this paper a new rule base reduction approach is suggested to manage large inference engines. Notions of rule hierarchy and sensor data fusion are introduced and combined to achieve desirable goals. New paradigms using soft computing approaches are utilized to design autonomous controllers for a number of robotic applications at the ACE Center are also presented briefly.

Introduction

Since the launching of Sputnik in the former Soviet Union, extensive progress has been achieved in our understanding of how to model, identify, represent, measure, control, and implement digital controllers for complex large-scale systems. However, to design systems having high MIQ® (Machine Intelligence Quotient), a profound change in the orientation of control theory may be required.

Currently, one of the more active areas of soft computing is fuzzy logic, and one of the more popular applications of fuzzy logic is fuzzy control systems. Fuzzy controllers are expert control systems that smoothly interpolate between hard-boundary crisp rules. Rules fire simultaneously to continuous degrees or strengths and the multiple resultant actions are combined into an interpolated result. Processing of uncertain information and saving of energy using common sense rules and natural language statements are the basis for fuzzy control. The use of sensor data in practical control systems involves several tasks that are usually done by a human in the decision loop, e.g., an astronaut adjusting the position of a satellite or putting it in the proper orbit, a driver adjusting a vehicle’s air-conditioning unit, etc. All such tasks must be performed based on the evaluation of data according to a set of rules which the human expert has learned from experience or training. Often, if not all the time, these rules are not crisp, i.e., some decisions are based on common sense or personal judgment. Such problems can be addressed by a set of fuzzy variables and rules which, if properly constructed, can make decisions as well as an expert.

This paper represents a set of applications of soft computing approaches to complex systems such as mobile robots, flexible arm, etc. The structure of the paper is as follows: Section 2 gives a brief introduction into autonomy through soft computing. Section 3 introduces two notions of sensory fusion and rule hierarchy. Section 4 constitutes a few applications of autonomous control for complex systems through soft computing approaches. Conclusions are described in Section 5.

Section snippets

Autonomy through soft computing

Soft computing is an umbrella terminology used to refer to a collection of intelligent approaches such as neural networks (NN) fuzzy logic (FL), genetic algorithm (GA), genetic programming (GP), etc. Soft computing techniques can allow the design of an autonomous controller through learning (NN), optimization (GA) or reasoning (FL).

Neural networks, GAs and genetic programming are augmented with fuzzy logic-based schemes to enhance artificial intelligence of automated systems. Such hybrid

Sensory fusion and rule hierarchy

In many real-life problems the number of sensory data is way too many for any reasonable sized rule base. For example, for a 4-variable system with five linguistic labels per variable, 625 rules are nominally needed. For a 10-variable process, the size of the rule base would be over 9.7 million. In other words, the size of the rule base would quickly approach infinity as the number of variables increases. In an effort to reduce the size of the rule base, many approaches are possible. Two of

Autonomous control in robotics

In this section many applications of soft computing towards rendering various degrees of autonomy in control systems will be presented. There are too much details in each of these case studies that we can cover in this short paper. Relevant references will help the readers in all case. Since 1993, the University of New Mexico’s CAD Laboratory and later on the ACE Center have been active in designing the autonomous behavior of many configurations of robots.

Conclusions

The basic theme of this paper was autonomous control and autonomy through several architectures of soft computing. A number of robotic applications were used to illustrate these architectures. These autonomous controllers are simple to implement in a laboratory environment on either a PC or on a chip-level board. Soon, autonomous control through intelligent paradigms technology will be a matter of economy and not controversy. It features applications in a wide variety of fields such as control,

Acknowledgements

This work was supported, in parts, by NASA Grant number NCCW-0087. The author wishes to thank three of his former students Dr. M. Akbarzadeh, Dr. K. Kumbla and Dr. E. Tunstel for their contributions.

References (10)

  • M.-R. Akbarzadeh, Fuzzy control and evolutionary optimization of complex systems, Ph.D. Dissertation, University of New...
  • M. Akbarzadeh, K. Kumbla, E. Tunstel Jr., M. Jamshidi, Soft computing for autonomous robotic systems, International...
  • D.E. Goldberg

    Genetic Algorithms in Search, Optimization and Machine Learning

    (1989)
  • A. Homaifar et al.

    Simultaneous design of membership functions and rule sets for fuzzy controllers using genetic algorithms

    IEEE Transactions on Fuzzy Systems

    (1995)
  • M. Jamshidi, Large-Scale Systems – Modeling, Control, and Fuzzy Logic, Prentice-Hall Series on Environmental and...
There are more references available in the full text version of this article.

Cited by (4)

  • Soft computing methods applied to the control of a flexible robot manipulator

    2009, Applied Soft Computing Journal
    Citation Excerpt :

    Soft computing, a collection of fuzzy logic technique, neural networks, evolutionary computation techniques has proven to be a powerful tool for adding autonomy to many complex systems [30].

  • Gas turbine diagnostics using a soft computing approach

    2006, Applied Mathematics and Computation
View full text