Saffiotti, A. (1997). The uses of fuzzy logic in autonomous robot navigation: a catalogue raisonn e. Soft Computing, 4(1):180--197.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....advantages and disadvantages [8, 4] Rather than merely giving a description of these methods, we highlight the main characteristics of multiple 1 Behaviors in this framework are equivalent to fuzzy behaviors because objective functions encode the semantics of fuzzy membership functions. See [10] for an excellent overview of application of fuzzy logic to behavior based robotics. 2 The action space is constrained due to various physical and geometrical properties, such as a set of non holonomic constraints imposed by the mechanical construction of the robot, into a subset of feasible ....

....so they can be interchanged as described in [9] Hence we can formulate the inferencing mechanism using a multiple objective decision theoretic approach, which provides formal tools for behavior coordination and resolving conflicts in a principled manner. Also with respect to behavior blending [10], fuzzy and multiple objective approaches are compatible and thus can be combined. One straightforward way to combine the two approaches is to: 1) Specify each behavior using a fuzzy rule base, and generate multivalued behavioral outputs using fuzzy inferencing, 2) Use fuzzy context rules to ....

Alessandro Saffiotti. The Uses of Fuzzy Logic in Autonomous Robot Navigation: a catalogue raisonn'e. Technical Report 2.1, IRIDIA, Universite Libr'e de Bruxelles, 50 av. F. Rossevelt, CP 194/6, B-1050 Brussels, Belgium, November 1997.

....robots from Real World Interface [Int] which includes several different methods. 2.1 Fuzzy methods In this section a couple of methods are presented which use fuzzy logic. For an overview of methods using fuzzy logic for mobile robots, we would like to refer to an overview written by Saffiotti [Saf97]. In figure 2.1 can be seen that a fuzzy set is capable of defining a lot of different kinds of uncertainty. Although just a few of them are used in the presented approaches, it opens a wide range of possibilities. As described in the previous chapter, a complete mobile robot control system at ....

A. Saffiotti. The uses of fuzzy logic in autonomous robot navigation: a catalogue raisonn'e. Soft Computing, 1(4):180--197, 1997. http://iridia.ulb.ac.be/saffiotti/papers.html.

....approach, which provides formal tools for behavior coordination and resolving conflicts in a principled manner. Preliminary experimental results are reported. 1 Introduction A significant body of work has demonstrated the advantages of fuzzy approaches to behavior based mobile robot control [10]. First, due to its approximate reasoning capabilities, fuzzy logic produces controllers that are robust to uncertainty (sensory noise, perturbations, etc. Further, fuzzy behaviors can be conveniently synthesized by a set of if then rules using easy to understand linguistic terms to encode ....

....fuzzy logic produces controllers that are robust to uncertainty (sensory noise, perturbations, etc. Further, fuzzy behaviors can be conveniently synthesized by a set of if then rules using easy to understand linguistic terms to encode expert knowledge. However, as pointed out by several authors [10], fuzzy behavior coordination suffers from problems associated with resolving behavioral conflicts. Fuzzy behavior coordination is performed by fuzzy inferencing methods (e.g. max min, prod sum) to combine the multivalued outputs of the behaviors and then defuzzification (e.g. centroid) is used ....

[Article contains additional citation context not shown here]

Alessandro Saffiotti. The Uses of Fuzzy Logic in Autonomous Robot Navigation: a catalogue raisonn'e. Technical Report 2.1, IRIDIA, Universite Libr'e de Bruxelles, 50 av. F. Rossevelt, CP 194/6, B-1050 Brussels, Belgium, November 1997.

....is based on a similar conceptual framework. However, as will be evident, its formalism affords certain advantages over fuzzy approaches. 3 Pitfalls of fuzzy behavior coordination A significant body of work has demonstrated the advantages of fuzzy approaches to behavior based mobile robot control [27]. Fuzzy methods describe a behavior (or its output) as a multivalued fuzzy membership function that reflects the (grade of) desirability of each action from that behavior s point of view. Fuzzy behaviors can conveniently be synthesized by a set of easy to understand fuzzy if then rules together ....

....by combining the fuzzy outputs of behaviors using an appropriate operator such as a triangular conorm (which corresponds to the fuzzy set union) e.g. the max operator. Then defuzzification (e.g. center of gravity, COG) is used to select a final crisp action ultimately used for control (see [27]) This scheme should lead to the selection of an action that somehow represents the consensus among the behaviors and thus best satisfies the decision objectives that they encode. This, however, is not always the case, due to certain shortcomings inherent in standard fuzzy inferencing mechanisms. ....

[Article contains additional citation context not shown here]

Alessandro Saffiotti. The Uses of Fuzzy Logic in Autonomous Robot Navigation: a catalogue raisonn'e. Technical Report 2.1, IRIDIA, Universite Libr'e de Bruxelles, November 1997.

....its superior an agent is viewed as an action. This is illustrated in figure 2. 5 A taxonomy for action selection mechanisms Classification of existing action selection mechanisms (ASMs) into a number of logical groups can be useful. In the review of the literature we find two taxonomies ( 32] and [56]) that suggest schemes for classification of ASMs. Agent Agent Agent Agent Agent Agent Agent agents actions Role wrt. subordinate action selection mechanism action Role wrt. superior agent Figure 2: Internal view of agencies. The two roles of an agent in the agent hierarchy. From its own point ....

....of sparse system resources. The authors of [32] further divide state based coordination mechanisms into temporal sequencing and competitive. Continuous State Based Competetive Temporal Sequencing Cooperative Coordination Classes Figure 3: Classes of coordination mechanisms proposed in [32] In [56] the ASMs are divided into arbitration and command fusion corresponding to MacKenzie s state based and continuous approaches respectively. In Saffiotti s terminology each of these classes are concerned with the problems of behavior coordination and command fusion respectively. Behavior ....

[Article contains additional citation context not shown here]

Alessandro Saffiotti. The Uses of Fuzzy Logic in Autonomous Robot Navigation: a catalogue raisonn'e. Technical Report 2.1, IRIDIA, Universite Libr'e de Bruxelles, 50 av. F. Rossevelt, CP 194/6, B-1050 Brussels, Belgium, November 1997.

....existing literature. Section 4 presents contextdependent blending, a general behavior coordination mechanism entirely grounded in fuzzy logic. Section 5 discusses the advantages of this mechanism. This note is part of a more comprehensive review of the uses of fuzzy logic in autonomous robotics [23]. 2. Behavior arbitration The arbitration policy determines which behavior(s) should influence the operation of the robot at each moment, and thus ultimately determines the task actually performed by the robot. Early behavior based architectures [8] relied on a fixed arbitration policy, ....

A. Saffiotti. The uses of fuzzy logic for autonomous robot navigation: a catalogue raisonn'e. Technical report, IRIDIA, Universit'e Libre de Bruxelles, Brussels, Belgium, 1997. Available on-line: http://iridia.ulb.ac.be/saffiotti/flarbib.html.

....uncertainty and imprecision; to build robust controllers starting from heuristic and qualitative models; and integrate symbolic reasoning and numeric computation in a natural framework. In the next pages, we shall illustrated these points by using our work on the robot Flakey as a test case. See [36] for an overview of the uses of fuzzy logic in autonomous robotics. 3 Robust behavior The first issue that we consider is the design of the individual behavior producing modules that appear in Figure 3. Each one of these modules fully implements a control policy for one specific sub task, or ....

....one of the three attitudes above and, if we opt for an explicit approach, by choosing a specific uncertainty formalism. Although most of the literature on dealing with uncertainty in robotics is based on probabilistic techniques, solutions based on fuzzy logic are being increasingly reported (see [36] for an overview) We have hinted at a few advantages of this choice in the pages above. Still, we emphasize that the choice of the formalism to use depends on the robot environment task configuration: there is no best way to deal with uncertainty in robotics, but there are as many best ways ....

A. Saffiotti. The uses of fuzzy logic for autonomous robot navigation: a catalogue raisonn'e. Technical report, IRIDIA, Universit'e Libre de Bruxelles, Brussels, Belgium, 1997. Available on-line: http://iridia.ulb.ac.be/saffiotti/flarbib.html.

No context found.

Saffiotti, A. (1997). The uses of fuzzy logic in autonomous robot navigation: a catalogue raisonn e. Soft Computing, 4(1):180--197.

No context found.

A. SAFFIOTTI, The uses of fuzzy logic in autonomous robot navigation: a catalogue raisonne. Soft Computing, vol. 1, no. 4, pp. 180--197, 1997.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC