A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering, School of Advanced Technologies

Thesis (M.Eng.) - Asian Institute of Technology, 2002

Time series are used in many fields for status variable monitoring; they give a picture of the behavior of a system with the time. The status variables variation reveal the dynamical behavior of a system, in the construction due to an important number of unpredictable factors and vague information (weather, site conditions, labor ... ), the dynamic of the system might seem to be totally random, and unexplainable. Then instead of understanding the dynamic of a system, the solution would be to model the vagueness of the information and compile them in a system that would be able to compute imprecise information. The fuzzy dynamical systems are based on such a perspective. Then a system based on these techniques is constructed for short time series generation of a precise status variable. Due to different scale integration for different factors the techniques exposed in this paper are limited to weekly plans. Nevertheless it is an efficient tool, in fact in the construction every working day is different and having a tool that integrated and gives the dynamical status of a system allows the manager or the scheduler to allocate efficiently the resources and schedule the degree of productivity regarding the objectives of the week and the reaction of the system under a given input.