Quasi-monotonic Segmentation of State Variable Behavior for Reactive Control

Real-world agents must react to changing conditions as they execute planned tasks. Conditions are typically monitored through time series representing state variables. While some predicates on these times series only consider one measure at a time, other predicates, sometimes called episodic predicates, consider sets of measures. We consider a special class of episodic predicates based on segmentation of the the measures into quasi-monotonic intervals where each interval is either quasi-increasing, quasi-decreasing, or quasi-flat. While being scale-based, this approach is also computational efficient and results can be computed exactly without need for approximation algorithms. Our approach is compared to linear spline and regression analysis.Les agents r\ue9els doivent r\ue9agir \ue0 des conditions variables quand ils ex\ue9cutent des t\ue2ches pr\ue9vues. Habituellement, les conditions sont surveill\ue9es via des s\ue9ries temporelles repr\ue9sentant des variables d'\ue9tat. M\ueame si certaines valeurs de pr\ue9vision de ces s\ue9ries temporelles ne prennent en consid\ue9ration qu'une mesure, d'autres, appel\ue9es valeurs de pr\ue9visions \ue9pisodiques, envisagent des ensembles de mesures. Nous consid\ue9rons une classe sp\ue9ciale de valeurs de pr\ue9vision \ue9pisodiques bas\ue9e sur la segmentation de mesures dans des intervalles quasi-monotones, o\uf9 chaque intervalle est soit quasi-croissante, quasi-d\ue9croissante ou quasi-uniforme. M\ueame si on tient compte de l'\ue9chelle, cette approche est aussi efficiente du point de vue calcul et les r\ue9sultats peuvent \ueatre calcul\ue9s exactement sans le recours \ue0 des algorithmes d'approximation. Notre approche est compar\ue9e \ue0 un spline lin\ue9aire et \ue0 l'analyse de r\ue9gression.NRC publication: Ye