Data Collection
Method where a sensor remains motionless for a fixed duration to average out random noise and determine a stable bias value provides a baseline for further measurements. The static dwell integration involves recording data at a high sample rate while the instrument is held in a precision fixture. By calculating the mean value over several minutes, the effect of short term fluctuations is reduced.
This technique is standard in the initialization of inertial navigation systems.
Averaging Period
Length of time the sensor spends in the stationary state determines the effectiveness of the noise reduction. Static dwell integration usually lasts from thirty seconds to several minutes depending on the stability of the device. A longer period allows the mathematical average to converge on the true bias.
However, excessively long dwells can introduce errors from thermal drift or changes in the environment. The optimal duration is found by analyzing the Allan variance of the sensor.
Noise Reduction
Statistical processing of the gathered data points lowers the standard deviation of the measurement. Static dwell integration filters out the white noise that is inherent in the electronic components of the sensor. This improvement in the signal to noise ratio allows the system to detect much smaller physical inputs.
The resulting bias estimate is then used to zero the instrument before a dynamic test begins. This procedure is repeated whenever the system is powered on.
Thermal Limit
Internal heating of the electronics during the stationary period can cause the bias to shift slightly. Static dwell integration must account for these changes by monitoring the temperature of the sensing element. If the temperature is not stable, the calculated average will not represent the true state of the sensor.
High precision units often include an internal heater to maintain a constant temperature during the dwell. Proper thermal management ensures the accuracy of the bias estimate.