Feedback linearization using CMAC neural networks

The CMAC has advantages over multilayer neural networks due to its increased structure which results in better computational speed and faster leaning. A repeatable design algorithm and stability proof is examined for an CMAC that uses basis-spline functions as receptive fields, unlike most standard adaptive control approaches which use basis vectors depending on the unknown plant (e.g. a tediously computed regression matrix). A ε-modification approach to adapt the CMAC parameters is investigated. With mild assumptions on the state-feedback linearizable nonlinear systems and using this CMAC, the uniform ultimate boundedness of the closed-loop signals is presented and that the controller achieves the desired tracking