Identification of Anomalies in Analysis of Hydroacoustic Measurements of Unsteady Processes

BACKGROUND: When processing hydroacoustic measurements, it is very important to identify and reject the so-called abnormal results as even one result can lead to false inference. The distinctive features of hydroacoustic measurements are, on the one hand, the high cost of errors in case of low accuracy and on the other hand, the difficulty of achieving the required accuracy due to the complexity and variability of the propagation medium and the background effect. The need to know the fine spatial and temporal structure of the measured hydroacoustic unsteady processes imposes additional strict requirements to the measurement methods and tools and their accuracy. The paper discusses and analyzes the issues related to the detection of anomalies in the analysis of hydroacoustic measurements of random non-stationary processes using the method of multiplication of estimates. AIM: To develop an algorithm for detecting abnormal measurements for stationary and non-stationary processes based on the method of multiplication of estimates. METHODS: To find the optimal parameter values of the proposed algorithms, we studied the relationship between the probability of correct detection and the probability of a false alarm and the number of multiplied estimates of the initial implementation and the number of covers per implementation length. It has been shown that in conditions of limited a priori data, this method allows to effectively determine the anomalous hydroacoustic measurements both for stationary and non-stationary processes. RESULTS: Using simulation modeling, we built models of measurements represented by a non-stationary random sequence. The studies have shown the high efficiency of detecting abnormal measurements of the proposed algorithm based on the method of multiplication of estimates. The high performance of the proposed algorithm in conditions of limited a priori data allows for its use in automated primary data processing systems.