In designing controllers for complex and ill defined nonlinear
dynamical systems there are needs not sufficiently addressed by
conventional control theory. These are mainly related to the problem
of environmental uncertainty and often call for human-like decision
making requiring the use of heuristic reasoning, learning from past
experience and a set of input-output crisp data describing the system.
In general, only one of the two information has been used in the
design phase: either a set of crisp input-output data or knowledge
acquired from experts. This research work deals with the development
of a new approach for implementing combining both information.
The simulation results show that the proposal effectively solve the
backing up a truck problem from several initial conditions presenting
a good robustness. The controller design and the simulation are
further presented and discussed.
