Some Observations on Stability of the Switching Adaptive Rule Bases of the Fuzzy Self-Adapted Interpolative Controllers

The fuzzy self adapted interpolative controllers are able to identify on-line and to control a wide range of unknown plants. These controllers are developed by combining fuzzy-inter-polative modules (implemented by look-up tables). In order to cope with contradictory strategies demanded by time variable plants, several adaptive correctors, corresponding to different operating regimes can be fusioned or switched. Switching adaptive rule bases can cause the same stability problems as switching simple controllers. A stability benchmark study is provided in this issue, the instability generating factors are revealed, and the use of the switching error is recommended.