Design of a dynamic inclinometer device and use of various methods to increase measurement stability

With advancements in MEMS technology, the usage of IMU (Inertial Measurement Unit) devices incorporating this technology has become widespread. In industrial applications, IMUs need to provide highly precise measurements depending on their area of use. While these precise measurements are made possible by the evolving IMU technology, some external factors can affect the measurement accuracy of IMUs. The most significant of these external factors are external acceleration and vibrations. The external acceleration and vibrations in the system to which the device is attached negatively impact the mechanical structure inside the IMU, leading to measurement errors. To mitigate these errors, the literature suggests using filters such as the Kalman filter, low-pass filter, etc. In this study, a dynamic tilt sensor device containing a 6-axis IMU has been designed and software-integrated with a Kalman filter, moving average filter, and low-pass filter. The performance of these filters within the designed tilt sensor under external acceleration conditions was analyzed in a computer environment using the prepared test platform and data recording system. In the analyses conducted using the RMSE method, the measurement error amounts of the sensor in degrees were compared. In the final stage of the study, a hybrid filtering method combining these three filters was designed. The hybrid filtering method is frequently used in literature. However, the novelty of this study lies in incorporating the designed moving average filter into the hybridization process along with the Kalman filter and low-pass filter. As a result of this specially designed hybrid filtering method within the scope of the study, average error values of 0.017° in static tests and 1.177° in dynamic tests were obtained. In light of the analysis results, it is seen that the hybrid filtering method in this study provides more successful or closer results compared to other hybrid filtering methods in the literature.