In this paper a model for planning with fuzzy resources is presented. Nearly all existing models of planning with resources need to work with exact value of the resources modified by actions execution. Therefore these models cannot deal with more realistic situations in which the resources quantities are not completely known. The model presented here allows to manage domains more tailored to real world, where preconditions and effects over resources can be specified in terms of uncertain intervals using fuzzy sets. Moreover it is possible to define problems with mixed logical/quantitative and pure numerical goals. The fuzzy approach adequately represents the kind of uncertainty existing in planning domains. The choice of trapezoidal fuzzy intervals represents a solution which combine a simple, but expressive representation, with computational simplicity. It is also presented a search strategy for goals over resources that provides a specific method to guide search. This strategy is necessary when solving “pure numerical problems”. Finally some experimental results are presented and commented.
Planning with fuzzy resources
BAIOLETTI, Marco;MILANI, Alfredo;POGGIONI, VALENTINA
2003
Abstract
In this paper a model for planning with fuzzy resources is presented. Nearly all existing models of planning with resources need to work with exact value of the resources modified by actions execution. Therefore these models cannot deal with more realistic situations in which the resources quantities are not completely known. The model presented here allows to manage domains more tailored to real world, where preconditions and effects over resources can be specified in terms of uncertain intervals using fuzzy sets. Moreover it is possible to define problems with mixed logical/quantitative and pure numerical goals. The fuzzy approach adequately represents the kind of uncertainty existing in planning domains. The choice of trapezoidal fuzzy intervals represents a solution which combine a simple, but expressive representation, with computational simplicity. It is also presented a search strategy for goals over resources that provides a specific method to guide search. This strategy is necessary when solving “pure numerical problems”. Finally some experimental results are presented and commented.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.