Monitoring of Bridge Malfunctions Based on Interval Estimates of the Variance of the Noise of Vibration Noisy Signals

The specific features of bridges are considered. It is noted that in the process of operation bridges are exposed to the influence of external factors, such as temperature changes, wind gusts, ice, heavy rain, seismic and landslide impacts, etc., which cause damage, corrosion, malfunction, etc. in bridge structures. Therefore, systems for monitoring the technical condition of bridges are developed. However, these monitoring systems do not allow to reveal the latent period of occurrence of a malfunction in the bridge structure and trace the dynamics of its development. It is noted that the formation of even the most insignificant faults in the bridge structures is accompanied by the emergence in the vibration signal of an additive noise correlated with the useful signal. Characteristics of this noise carry information about the formation of defects and malfunctions in the bridge structures. However, the noise cannot be isolated from the noisy vibration signal in order to calculate estimates of these characteristics using traditional methods. Therefore, algorithms for calculating the characteristics of the noise characteristics of the noisy vibration signal are developed. Technologies for early detection of bridge malfunctions are proposed, as well as analysis of the dynamics of their development using interval estimates of the noise variance. The matrix and model of the faulty technical condition of bridges are built. To control the dynamics of malfunction development, confidence intervals for the estimates of the variance of the noise of noisy vibration signal at different time instants are compiled. The correspondence between each value of the confidence interval of the variance of the noise of the noisy vibration signal and the degree of development of the malfunction or defect of the bridge structure is established. For each upper boundary of the confidence interval of the noise variance, the ratio of the noise variance to the variance of the useful signal is calculated, and the dynamics of the hazard degree development is determined. It is shown that using the developed algorithms and technologies in monitoring systems will make it possible to signal the emergence of malfunctions and defects in the latent period of their inception, which ensures safety, reliability and efficiency of the operation of bridge structures and reduces the risk of possible accidents.