| Summary: | 碩士 === 逢甲大學 === 電子工程所 === 97 === During the past years, the integrated circuits electromagnetic compatibility (ICEMC) has permanently gained more and more significance. At present, ICs are moving towards operation frequencies in the GHz range with rise and fall times in the order of a few pico seconds. While the frequency increase, parasitic components of the package and silicon include impedances, would produce electromagnetic interference for circuit designer and cause malfunction, therefore, the designer should construct the model of integrated circuits to predict the electromagnetic interference.
In this thesis, we propose both the parameter estimation of IC-EMC model by fuzzy logic system and construct the equivalent model by fuzzy neural network. There two structures of IC-EMC model are included: a pair of symmetric parasitic capacitors added in the ICEM model and parameter change in the pair of symmetric ICEM model. Furthermore, two behavioral models of fuzzy neural network are constructed: type-1 fuzzy neural network and type-2 fuzzy neural network. The simulation results confirm the validity and performance of the advocated technique.
|
|---|
