Theoretical aspects of the CMAC and its application to high-dimensional aerospace modelling problems

The structured cerebellar model articulation controller (CMAC) offers a model-based, adaptive, but more reliable online neuro-control for the aerospace industry. The curse of dimensionality can be circumvented by exploiting any redundancy in the underlying function. Active structural load alleviating flight control systems are one such area where this work has shown that the controllers may be modelled using suitable CMAC network architecture and deployment strategies. This work builds a set of baseline rules which are to be used in the training ends synthesis of CMAC based adaptive neuro-controllers. This is proposed as a substitute for the present more traditional combination of gain-scheduled, plus model reference, MIT rule, Lyapunov, or parameter adaption schemes currently used in control for the aerospace industry.