Some studies on uncertainty management in dynamical systems using cybernetic approaches and fuzzy techniques with applications

  • Authors:
  • Kausik Kumar Majumdar;D. Dutta Majumder

  • Affiliations:
  • Electronics and Communication Sciences Unit, Indian Statistical Institute, 203, B.T. Road, Calcutta-108, India;Electronics and Communication Sciences Unit, Indian Statistical Institute, 203, B.T. Road, Calcutta-108, India

  • Venue:
  • International Journal of Systems Science
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

Uncertain information processing by fuzzy if-then rules has received much attention. Here we have taken a different path to model a system, about which we do not have precise information, namely modelling the system by fuzzy-valued functions without resorting to fuzzy if-then rules. As a result, the phase (state) space of the system becomes a fuzzy set and the underlying fuzzy mapping becomes a fuzzy attainability set mapping. Uncertain or fuzzy dynamical systems have been defined in terms of fuzzy attainability set mappings. Fuzzy differentiable dynamical systems have been discussed with a particular emphasis on fuzzy differential inclusion (FDI) relations. An evolutionary algorithm for solving one-dimensional FDIs has been developed. A model of the creation of a tropical cyclone in the form of a vortex, created by winds coming from different directions and colliding under certain conditions, has been proposed in terms of FDIs. A model has been considered for the highly uncertain system of evolution of a tumour in a human body and an FDI relation model of the whole system has been proposed and simulated. The model of an evolution of turbulence, as a random occurrence of vortices in a three-dimensional dynamic fluid, has been proposed and simulated, where each vortex is modelled by a fuzzy-valued function, where uncertain parameter and variable values are fuzzy numbers. All the systems represented by FDI relations have been simulated with the help of the evolutionary algorithm mentioned above.