Biosocial Culture Inspired Hierarchical Algorithm for MISO Block Oriented Nonlinear System Identific

Abstract A new solution to nonlinear systems identification of MISO Hammerstein and/or Wiener models is developed using tools from Evolutionary Computation based optimization. A hybrid genetic and swarming intelligence based Hierarchical Cultural Algorithm is used to adapt the structure of the bad less suited model and to estimate the parameters of its dynamics and nonlinearities representation. Performances of such solution, illustrated through a real life application show how this class of tools can be very helpful to solve complex nonlinear problems such as the ozone phenomenon identification.