Reference Number: INTAS-97-0606
INTAS logo Development of an Intelligent Sensing Instrumentation Structure I.S.I.S.
Using Neural Networks in ISIS
Home | The main features | Contradictions | Simulating of drifts | Combination of drifts | Error correction
The main features of accuracy increasing methods
In modern measurement systems the measurement error is defined by sensor. It is caused in significant initial spread of the sensor transformation characteristics and sensor drift during exploitation. The constructing-technology methods of sensor accuracy improving are reach therefore the structure-algorithmic methods are more preferable. Among last ones the calibration and prediction methods are most suitable. The main features of these methods are presented in Figure 1.

Figure 1. The main features of accuracy increasing methods
The correction by formulas requires the real parameters of these formulas with allowable accuracy. Using of the standardised formulas is possible in separate cases, for example at linearization. Using of the individual formulas is possible for correction of sensor drift. But pure prediction isn't reliable and requires many previous researches. Therefore the most optimal is combination of calibration or testing methods and prediction method, where prediction allows to increase intercalibration interval. The increasing of intercalibration interval is possible at high quality prediction and artificial neural networks provide the best quality of prediction of complex functions of sensor drift.