Many different sensing principles for inclination measurement are commonly used ([1], [2]). A well known approach is to track movements of a pendulum caused by the force of gravity with conductive, resistive or capacitive measurement systems ([3], [4]). However, with reasonable dimensions of the sensor its accuracy is limited to about +-0.1 deg due to unavoidable mechanical hysteresis even with very precise mechanical construction. Using a fluid instead of a mechanical pendulum together with a capacitive sensing principle helps to overcome some drawbacks of the mechanical pendulum while providing better accuracy and mechanical stability due to its insensitivity to shocks and vibrations. However, guaranteeing an adequate dynamic behavior of the sensor is much more difficult due to the less predictable nature of the fluid. Focusing on the usability of a liquid based sensor as part of an overload protection system for mobile cranes, this would lead to higher security margins. With the use of an adaptive signal processing strategies presented in this paper, undesired dynamic effects caused by the fluid can be significantly lowered. Furthermore, the reliability information gathered by the filter may be used for self diagnostics allowing the sensor to identify inconsistencies in the measured data with very little additional computational costs. In addition, fault tolerance may be achieved by supplementing the adaptive approach by a reconstruction algorithm which allows for the recovery of corrupted measurement data. This leads to a robust smart sensor system.
展开▼