An inexpensive and yet good method for measuring inclination is to use so-called MEMS sensors. These consist of electronic and micromechanical structures, so that in addition to the usual microelectronic circuits, there are also mechanically movable structures in the chips. These structures deflect when the earth's g-force acts on them. At the same time, the capacitances (capacitors) formed by these structures change, which are measured in order to calculate and output an angle of inclination.

The combined sensors with an operating temperature range between -40 and +85° C measure inclination angles of up to ±100° with a resolution of 0.01° and an accuracy of ±0.1°. Measurements are possible in one or two axes.

© TWK Electronics

However, if this MEMS cell is exposed to accelerations other than the acceleration due to gravity - for example due to vibrations or acceleration in a vehicle - the MEMS structures are also moved and a falsified inclination value is output. As a result, the connected control system may react incorrectly and an accident occurs, for example with mobile machines such as mobile cranes, concrete pumps or forklift trucks.

This is where TWK brings a gyroscope into play - a rotation rate sensor also designed using MEMS technology. Its micromechanical structures are designed in such a way that angular velocities (°/s) are measured and output as an output signal. The gyroscope therefore measures changes in the angle of inclination and not the angle of inclination itself when the MEMS gyroscope rotates around its own axis or an extended axis. The linear accelerations mentioned above, which disturb the acceleration MEMS sensor, do not affect the gyroscope MEMS sensor as they do not represent a rotation.

But how can the two sensors be combined to form a tilt measurement system that is insensitive to interference?

Optimizing the system

Measurement of the change in inclination with disturbance components (red). Averaging over 1000 values (green) smoothes out moderate vibrations, but causes a time delay. Blue shows the merging of the sensors via Kalman filter.

© TWK Electronics

The solution is to combine both measurement signals and optimize them using a Kalman filter. This filter technology makes it possible to calculate, correct and predict the measured value to be output, i.e. to determine it in advance for a short period of time. This is necessary because the acceleration sensor cannot provide an exact measured inclination value in the event of dynamic disturbance accelerations. Instead, a recursive algorithm is used to estimate the next output values based on the previous measured values and constantly optimize them in real time (cycle time <1 ms) with the help of new real measured values. The signal from the angular rate sensor must first be integrated over time so that it can be added. In addition, a number of parameters are used to optimize the system as much as possible, to smooth the measured values and, if necessary, to adapt them to application-specific conditions. Application-related simulations and measurements at TWK have shown that the sensor fusion system works precisely and reliably and enables update rates of significantly less than 1 ms.

Insensitive to vibrations

The sensor system ensures safe, hazard-avoiding use of mobile machinery, even during the operation of motors, hydraulic pumps or other actions that generate strong vibrations. It is also largely insensitive to disruptive acceleration and lateral tilt deflection.

One example of the use of fused sensors is mobile concrete pumps. These vehicles drive to a construction site, are aligned - i.e. the undercarriage is set horizontally and the boom is extended with the trunk. The undercarriage and boom are fitted with inclination sensors. When the concrete pump delivers the concrete, short but strong impacts occur. The resulting accelerations interfere with the inclination measurement. However, the signal from the gyroscope is 'zero' despite the linear interference acceleration. The calculation algorithm then recognizes that there is no change in inclination and retains the previous inclination value, which is now subject to strong fluctuations on the MEMS side. If the inclination changes exactly at the time of the disturbance, the gyroscope outputs a change in °/s, the integral of which is not 'zero'. Both values are considered by the Kalman filter and in turn used to generate a usable inclination value.ik

Author:
Achim Albertini is Product Manager at TWK-Elektronik in Düsseldorf.