A detailed systematic review of the constraint satisfaction literature.
Methods are developed which, for the first time, are able to solve problems which both contain a dynamic component and are open to compromise if a ‘perfect’ solution does not exist Classical artificial intelligence planning is extended to incorporate preferences so that it too can support compromise A trade-off between the length of a plan versus the number and severity of the compromises it contains is now possible An extensive empirical analysis of the new dynamic-flexible problem solving methods and the development of a new flexible planning Includes supplementary material: sn.pub/extras
Ian Miguel
Constraint Satisfaction Extension algorithms artificial intelligence complexity constraint satisfaction problem fuzzy heuristics intelligence knowledge knowledge representation logistics