Modeling and control of unknown chaotic systems via multiple models

This paper presents a multiple model based control methodology with sliding manifold for unknown chaotic systems from a finite number of measurements of the inputs and outputs. The self-organizing map (SOM) is employed to quantize the operating regions such that a local linear model is built for each region in which a sliding mode controller is designed based on the model. Switching of the controllers is done synchronously with the active local linear model that tracks the different operating conditions. It is the purpose of this work to regulate the unknown chaos to a fixed point or a stable periodic orbit. Simulations on the unknown controlled Lorenz system illustrate the efficiency of the proposed control technique