Kombinierter Einsatz von Gassensoren verschiedener Technologien (KombiSens). Teilprojekt: Fuzzy-Auswertung in der Gassensorik Abschlussbericht

Substrate Temperature Control with Fuzzy Logic: On-chip integrable gas sensors show temperature dependent characteristics, forcing users to keep the temperature of the sensors at a constant level, but also allowing increased accurcy and selectivity of gas classification by deliberate temperature changes. Problems with modelling and integrability are easier to solve by implementation of a fuzzy controller, but there remains a substantial hardware effort on the chip. As a result the layout of a chip was developed with areas whose temperature can be independently controlled. These areas can be utilized by any user to stabilize temperature dependent circuits or to test their behaviour at varying temperatures. Combined Neuro/Fuzzy Signal Processing: In applications of gas sensors their output signals depend on parasitic influences like air humidity, temperature deviations or the age of the sensor. The development of a gas classifier based on statistical methods involves through the necessary number of samples considerable time and cost. With fuzzy methods a gas classifier can be easier constructed; additional results, like p.e. the reliability of the classification, may be used in safety relevent applications. Fuzzy logic is equally suited to mutual monitoring of gas sensors in a multiple sensor system. A functional working model of such a fuzzy state identification system was completed. (orig.)SIGLEAvailable from TIB Hannover: F97B571+a / FIZ - Fachinformationszzentrum Karlsruhe / TIB - Technische InformationsbibliothekBundesministerium fuer Bildung, Wissenschaft, Forschung und Technologie, Bonn (Germany)DEGerman