Hi friends! There are literature pertaining to design of adaptive fuzzy controller most notably by Wuxi Shi that delves deeper into design of adaptive fuzzy controller for MIMO system with nonsymmetric gain and unknown gain direction, however for my problem I quickly realized that the issue stems from the formulation of robustifying term- ur. It is quickly becoming NaN.

Similarly in Tong's work I realized that the performance of the controller is heavily dependent on initialization of the parameters. For some initial values of the parameters, the controller was able to drive the error to zero, however for certain others, it couldnt. This issue was solved by W. Shi using a robusifying term containing Nussbaum- however as pointed out, this robusifying term is getting NaN for my problem.

Kindly suggest me ways to overcome this problem or literature that has solved this issue satisfactorily. Thanks in advanc