Sensors for the detection of gaseous components are based on a variety of operating principles and materials. Several of them have found widespread application. For example, semiconductor gas sensors using metal oxides, which can detect explosive gases in air such as hydrogen, hydrocarbons or organic vapours, are used for gas leakage alarms in quantities of some million pieces. The global issues of energy and environment evoked a demand for sensors which respond to air pollutants in environments such as carbon oxides or halogenated hydrocarbons. The application of artificial neural networks (ANN) in gas sensor signal evaluation can serve different purposes and can replace conventional methods: - the characteristics of the sensor elements can be represented by a polynomial fit, a look-up table or by an artificial neural network - in the case of slow reacting sensor elements the reading can be speeded up by modelling the response in a differential equation (e.g. for linear responding electrochemical cells by a linear equation of 2nd order). A model-free access is possible by a neural network - compensation of cross sensitivities in multisensors
Testability of gas sensors in neural networks
Testbarkeit von Gassensoren in neuronalen Netzen
1995
7 Seiten, 6 Bilder
Aufsatz (Konferenz)
Englisch
Testability of micromachined silicon rate sensors
AIAA | 1997
|Design of vehicle networks for testability
Kraftfahrwesen | 1990
|Design of vehicle networks for testability
Kraftfahrwesen | 1989
|Design of vehicle networks for testability
Tema Archiv | 1988
|Predicting testability of program modules using a neural network
Tema Archiv | 2000
|