Quality comfirmation of two typical acceleration data based on Benford's law

Abstract

In order to truly explore the health status of bridge structures, it is usually necessary to use bridge health monitoring system for long-term monitoring. Acceleration sensor is widely used in health monitoring system. However, on one hand, the acceleration sensor often has unstable signal and noise due to electromagnetic interference, cable damage and other reasons in complex outdoor environment. On the other hand, there is a lack of fast and effective data quality evaluation methods before using data. Therefore, this paper proposes a data quality evaluation method based on Benford’ s law, and applies this method to two typical acceleration response data cases for data analysis. The data of the two cases are respectively taken from the acceleration response at the cable of a cable-stayed bridge and the finite element simulation acceleration signal of a bridge sound barrier under the influence of wind load. Firstly, the digit statistical algorithm is used to count the distribution of the digit of the testing data. Secondly, the data is processed by filtering signal or adding noise to study the influence of noise on the quality of acceleration data. Finally, the statistical results were tested and the test values of each group were compared and analyzed. The research results prove that the data quality of current electronic sensors is reduced under the influence of noise. The research content of this paper not only provides a new idea for structural health monitoring acceleration data, but also this method is expected to be applied to other data of bridge health monitoring. Related research has great development prospect and scientific research value in the field of verification of sensor data.