Enrique H. Ruspini

A multivalued logic approach to integrating planning and control

Abstract elligent agents embedded in a dynamic, uncertain environment should incorporate capabilities for both planned and reactive behavior. Many current solutions to this dual need focus on one aspect, and treat the other one as secondary. We propose an approach for integrating planning and control based on behavior schemas , which link physical movements to abstract action descriptions. Behavior schemas describe behaviors of an agent, expressed as trajectories of control actions in an environment, and goals can be defined as predicates on these trajectories. Goals and behaviors can be combined to produce conjoint goals and complex controls. The ability of multivalued logics to represent graded preferences allows us to formulate tradeoffs in the combination. Two composition theorems relate complex controls to complex goals, and provide the key to using standard knowledge-based deliberation techniques to generate complex controllers. We report experiments in planning and execution on a mobile robot platform, Flakey.

On the semantics of fuzzy logic

Abstract This paper presents a formal characterization of the major concepts and constructs of fuzzy logic in terms of notions of distance, closeness, and similarity between pairs of possible worlds. The formalism is a direct extension (by recognition of multiple degrees of accessibility, conceivability, or reachability) of the najor modal logic concepts of possible and necessary truth. Given a function that maps pairs of possible worlds into a number between 0 and 1, generalizing the conventional concept of an equivalence relation, the major constructs of fuzzy logic (conditional and unconditioned possibility distributions) are defined in terms of this similarity relation using familiar concepts from the mathematical theory of metric spaces. This interpretation is different in nature and character from the typical, chance-oriented, meanings associated with probabilistic concepts, which are grounded on the mathematical notion of set measure. The similarity structure defines a topological notion of continuity in the space of possible worlds (and in that of its subsets, i.e., propositions) that allows a form of logical “extrapolation” between possible worlds. This logical extrapolation operation corresponds to the major deductive rule of fuzzy logic — the compositional rule of inference or generalized modus ponens of Zadeh — an inferential operation that generalizes its classical counterpart by virtue of its ability to be utilized when propositions representing available evidence match only approximately the antecedents of conditional propositions. The relations between the similarity-based interpretation of the role of conditional possibility distributions and the approximate inferential procedures of Baldwin are also discussed. A straightforward extension of the theory to the case where the similarity scale is symbolic rather than numeric is described. The problem of generating similarity functions from a given set of possibility distributions, with the latter interpreted as defining a number of (graded) discernibility relations and the former as the result of combining them into a joint measure of distinguishability between possible worlds, is briefly discussed.

The paradoxical success of fuzzy logic

Fuzzy logic methods have been used successfully in many real-world applications, but the foundations of fuzzy logic remain under attack. Taken together, these two facts constitute a paradox. A second paradox is that almost all of the successful fuzzy logic applications are embedded controllers, while most of the theoretical papers on fuzzy methods deal with knowledge representation and reasoning. I hope to resolve these paradoxes by identifying which aspects of fuzzy logic render it useful in practice, and which aspects are inessential. My conclusions are based on a mathematical result, on a survey of literature on the use of fuzzy logic in heuristic control and in expert systems, and on practical experience in developing expert systems.<<ETX>>

The Saphira architecture: a design for autonomy

Mobile robots, if they are to perform useful tasks and become accepted in open environments, must be fully autonomous. Autonomy has many different aspects ; here the focus is on three central ones: the ability to attend to another agent, to take advice about the environment, and to carry out assigned tasks. All three involve complex sensing and planning operations on the part of the robot, including the use of visual tracking of humans, coordination of motor controls, and planning. It is shown how these capabilities are integrated in the Saphira architecture, using the concepts of coordination of behaviour, coherence of modelling, and communication with other agents.

Handbook of Fuzzy Computation

Foreword by Lotfi Zadeh Preface INTRODUCTION Background. Why fuzzy logic? FUNDAMENTAL CONCEPTS OF FUZZY COMPUTATION Vagueness and uncertainty: Theories of vagueness. Theories of uncertainty. Fuzzy sets: concepts and characterizations: Introduction. Operations on fuzzy sets. Interpretations of fuzzy sets. Fuzzy relations. Characterization of fuzzy sets. Fuzzy measure and integral. Fuzzy mathematical objects. Extension principle. Fuzzy set calculus: Introduction. Membership function elicitation. Fuzzy relational calculus. Fuzzy arithmetic. Possibility theory. Fuzzy reasoning: Introduction. Fuzzy inference. Defuzzification. FUZZY MODELS Fuzzy models. Modeling and simulation: Granule-based models. Logical aspects of fuzzy models. Statistical models. Fuzzy Petri Net model. Model acquisition. Approximation aspects of fuzzy models. HYBRID APPROACHES Introduction: motivation for hybrid approaches. Neuro-fuzzy systems. Fuzzy-evolutionary systems. FUZZY COMPUTATION ENVIRONMENTS Software approaches: Programming languages. Knowledge-based systems. Database management, information retrieval, and decision support systems. Hardware approaches: Desirable features. Adapting existing hardware to fuzzy computation. Analog approaches. Digital approaches. Hybrid (digital-analog) approaches. APPLICATIONS OF FUZZY COMPUTATION Knowledge based systems: Knowledge representation. Inference methods. Control methods. Design methods. Control. Principles of fuzzy controllers. Fuzzy control approaches: General design schemes. Cell maps. Sliding mode control. Predictive control. Hierarchical control. Model-based control. Optimal fuzzy control. Machine learning: Introduction: learning fuzzy concepts. Supervised learning. Reinforcement learning. Data and information management: Fuzzy databases. Information retrieval. Case-based reasoning. Decision making and optimization: Decision-making models. Optimization. Pattern analysis. Computer vision. FUZZY COMPUTATION IN PRACTICE Aerospace: Proximity operations spacecraft controller: a case study in fuzzy logic control. Systems control: DC/DC converters fuzzy control. Fuzzy control in telecommunications. Fuzzy-neural traffic control and forecasting. Systems control. Backlash compensation using fuzzy logic. Neurofuzzy modeling for nonlinear system identification. Nuclear engineering: Application of fuzzy logic control system for nuclear reactor control. Manufacturing: Applications of fuzzy set methodologies in manufacturing. Compensation of friction in mechanical positioning systems. Diagnostics: Possibilistic handling of uncertainty in fault diagnosis. Robotics: Autonomous mobile robot control. Chemical engineering: Chemical engineering application. Water treatment: Water treatment application. Automotive: Improvement of the relationship between driver and vehicle using fuzzy logic. Traffic engineering: Traffic engineering application. Civil engineering: Civil engineering application. Engineering design: A fuzzy sets application to preliminary passenger vehicle structure design. Oil refining: Neuro-fuzzy hybrid control system in petroleum plant. Medicine: CADIAG2: hospital-based computer-assisted differential diagnosis in internal medicine. Neural networks for ECG diagnostic classification. Information science: Case-based reasoning. Information retrieval: a case study of the CASHE: PVS systems. Economics, finance and business. Decision support system for foreign exchange trade (FOREX). Operations research: Scheduling. Fuzzy sets in operation research: forecasting, a case study. Quality design using possibilistic regression and optimization. Inventory control. Time series prediction. FUZZY COMPUTATION RESEARCH Directions for future research. APPENDICES

Blending reactivity and goal-directedness in a fuzzy controller

Controlling the movement of an autonomous mobile robot requires the ability to pursue strategic goals in a highly reactive way. The authors describe a fuzzy controller for such a mobile robot that can take abstract goals into consideration. Through the use of fuzzy logic, reactive behavior, e.g., avoiding obstacles on the way, and goal-oriented behavior, e.g., trying to reach a given location, are smoothly blended into one sequence of control actions. The technique proposed has been implemented in the SRI mobile robot Flakey, resulting in extremely smooth and reliable movement.<<ETX>>

Using fuzzy logic for mobile robot control

The development of techniques for autonomous operation in real-world, unstructured environments constitutes one of the major trends in the current research on mobile robotics. In spite of recent advances, a number of fundamental difficulties remain. In this chapter, we discuss how fuzzy logic techniques can be used to address some of these difficulties. To illustrate the discussion, we describe the fuzzy-logic solutions developed on Flakey, the mobile robot of SRI International.

Epistemic logics, probability, and the calculus of evidence

This paper presents results of the application to epistemic logic structures of the method proposed by Carnap for the development of logical foundations of probability theory. These results, which provide firm conceptual bases for the Dempster-Shafer calculus of evidence, are derived by exclusively using basic concepts from probability and modal logic theories, without resorting to any other theoretical notions or structures. n nA form of epistemic logic (equivalent in power to the modal system S5), is used to define a space of possible worlds or states of affairs. This space, called the epistemic universe, consists of all possible combined descriptions of the state of the real world and of the state of knowledge that certain rational agents have about it. These representations generalize those derived by Carnap, which were confined exclusively to descriptions of possible states of the real world. n nProbabilities defined on certain classes of sets of this universe, representing different states of knowledge about the world, have the properties of the major functions of the Dempster-Shafer calculus of evidence: belief functions and mass assignments. The importance of these epistemic probabilities lies in their ability to represent the effect of uncertain evidence in the states of knowledge of rational agents. Furthermore, if an epistemic probability is extended to a probability function defined over subsets of the epistemic universe that represent true states of the real world, then any such extension must satisfy the well-known interval bounds derived from the Dempster-Shafer theory. n nApplication of this logic-based approach to problems of knowledge integration results in a general expression, called the additive combination formula, which can be applied to a wide variety of problems of integration of dependent and independent knowledge. Under assumptions of probabilistic independence this formula is equivalent to Dempsters rule of combination.

Issues and challenges of knowledge representation and reasoning methods in situation assessment (Level 2 Fusion)

Situation assessment (SA) involves deriving relations among entities, e.g., the aggregation of object states (i.e. classification and location). While SA has been recognized in the information fusion and human factors literature, there still exist open questions regarding knowledge representation and reasoning methods to afford SA. For instance, while lots of data is collected over a region of interest, how does this information get presented to an attention constrained user? The information overload can deteriorate cognitive reasoning so a pragmatic solution to knowledge representation is needed for effective and efficient situation understanding. In this paper, we present issues associated with Level 2 (Situation Assessment) including: (1) user perception and perceptual reasoning representation, (2) knowledge discovery process models, (3) procedural versus logical reasoning about relationships, (4) user-fusion interaction through performance metrics, and (5) syntactic and semantic representations. While a definitive conclusion is not the aim of the paper, many critical issues are proposed in order to characterize future successful strategies to knowledge representation and reasoning strategies for situation assessment.

Approximate reasoning: past, present, future

Abstract This paper presents a personal view of the state of the art in the representation and manipulation of imprecise and uncertain information by automated processing systems. To contrast their objectives and characteristics with the sound deductive procedures of classical logic, methodologies developed for that purpose are usually described as relying on approximate reasoning. Using a unified descriptive framework, we will argue that, far from being mere approximations of logically correct procedures, approximate reasoning methods are also sound techniques that describe the properties of a set of conceivable states of a real-world system. This framework, which is based on the logical notion of possible worlds, permits the description of the various approximate reasoning methods and techniques and simplifies their comparison. More importantly, our descriptive model facilitates the understanding of the fundamental conceptual characteristics of the major methodologies. We examine first the development of approximate reasoning methods from early advances to the present state of the art, commenting also on the technical motivation for the introduction of certain controversial approaches. Our unifying semantic model is then introduced to explain the formal concepts and structures of the major approximate reasoning methodologies: classical probability calculus, the Dempster-Shafer calculus of evidence, and fuzzy (possibilistic) logic. In particular, we discuss the basic conceptual differences between probabilistic and possibilistic approaches. Finally, we take a critical look at the controversy about the need and utility for diverse methodologies, and assess requirements for future research and development.