Error Correction
Computational correction algorithms use data from secondary transducers to remove systemic errors from a primary measurement signal. Multi sensor compensation is necessary because no single sensing element is perfectly immune to environmental interference like temperature or pressure. By measuring these interfering variables simultaneously, a processor can apply a mathematical model to nullify their effects.
This results in a much higher level of accuracy than a standalone sensor could achieve. It is standard practice in high precision inertial navigation units.
Cross Sensitivity
Primary sensors often react to forces they are not designed to measure. For instance, a pressure sensor might change its output when the temperature rises. Multi sensor compensation uses a dedicated temperature probe to identify this shift and subtract it from the final reading.
Polynomial Mapping
Mathematical models used for this correction are often based on a multi-dimensional surface fit. During the factory calibration, multi sensor compensation parameters are calculated by exposing the device to a matrix of known conditions. These coefficients are then stored in the non-volatile memory of the sensor.
Processing Overhead
Latency in the output can occur if the correction calculations are too complex. The multi sensor compensation algorithm must be optimized for the speed of the local microcontroller. This ensures that the corrected data is available in real time for control loops.
