MATHEMATICAL MODEL FOR AUTOMATED SYSTEMS OF SEISMOACOUSTIC MONITORING OF A MORTAR EXPLOSION CLASSIFICATION FOR CONDUCTING REMOTE RECONNAISSANCE

Abstract

Reconnaissance is the most important type of ensuring military operations; it is a set of measures taken by all commanders and staff to timely obtain information about the enemy, the terrain, climatic and weather conditions in the area of future hostilities to most effectively use their forces and means to defeat the enemy. The paper presents mathematical models for flavored seismo-acoustic monitoring systems for evaluating classes of mortars that fire to conduct remote reconnaissance. When evaluating the parameters of the mathematical model for seismoacoustic monitoring systems of the mortar explosion signal in the general case, we are faced with such a presentation of the seismoacoustic field model, which is observed when observations are complicated by additive interference. The model depends on the temporal and spatial coordinates and the vector of informative parameters of the model characterizing the research object. The model of the field formation process is defined by two vectors of free parameters of the model, and the model itself is the researcher’s hypothesis about the process being modeled. We divide the parameter vectors according to how they enter the model, linearly and non-linearly. Object classification processes aim to investigate the differences between signals from different mortars to conduct remote reconnaissance for the classification of the object under study. The paper presents mathematical models of automated seismo-acoustic monitoring systems for conducting remote reconnaissance for mortar classification based on seismo-acoustic records of artillery fire. In this work, detection processes operating on one recording channel are considered. With the help of the proposed model, the object under investigation is mapped into the vector of informative parameters of the model, which characterizes this type of object. Having collected statistics on a set of different types of objects in various conditions, we can build a procedure for classifying the objects under investigation based on the classification of the vector of informative parameters of the model. The paper considers the generalization of the mathematical model of a single explosion in the case of overlapping signals in the process of recording the wave seismic field. Mathematical models of automated seismo-acoustic monitoring systems are used to model fields of mechanical elastic waves. This work presents such a model for identifying mortar weapons for remote reconnaissance. It can be concluded that this parametric model reflects the process in the feature space and characterizes the object that fires shots. Thus, the presented model maps each type of mortar fire into its n-dimensional vector of informative parameters, which enables the classification of small arms. Therefore, an effective analysis method for estimating the parameters of mortar explosion signals and an unconventional model of the natural background on which mortar explosion signals are recorded is proposed.