الفهرس يوجد فقط 14 صفحة متاحة للعرض العام
 |  | from | 140 |  |  |
| | | from | 140 | | |
المستخلص
This thesis proposes an adaptation algorithm for adapting fuzzy controller parameters; namely, the Rule Degree Adaptation algorithm (RDA). This algorithrn-is introduced to minimize the controller output error and the plant output error.
Most existing methods of fuzzy adaptive control employ some kind of plant model, which is used to deduce the error of the control signal from either the error in the plant output or the error in the controller output. The error is used to adjust the controller parameters such that some cost function is minimized. The controller readjusted parameters may be the parameters of the membership function,’ input /output scale factor, rules or any combination of these parameters, assuming that the degree of fulfillment for all rules in the rule base is the same.
Unlike these, in the rule degree adaptation algorithm, each rule is assumed to possess its own degree of validity and the algorithm deduces the suitable degree of validity for each rule in order to satisfy the required performance index.
It is proven that by the use of mathematical analysis and simulation experiments, the control signal and the output of the plant being controlled will be both bounded and convergent. Experiments demonstrate that the algorithm yields performance, which is comparable or superior to that yielded by other adaptive fuzzy algorithms, conventional (PID) controller, and fixed parameters fuzzy paradigms. In addition to these results, this thesis shows the following:
• The (RDA) algorithm may be expected to work as well on reference controller output as it does on reference plant output. Furthermore, the algorithm can be applied on fast dynamic systems as well as slow systems.
• The effect of the disturbance on the system response for different adaptation algorithms shows the efficiency of the proposed method.
• The effect of the adaptation factor on the system stability in the case of reference controller output and reference plant output for (RDA) adaptation algorithm is also considered.