A Support Vector Machine-assisted Underwater Maneuvering Target Tracking Method Based on Interactive Multi-Model

Abstract

The invention proposes an interactive multi-model-based support vector machine-assisted underwater maneuvering target tracking method, which belongs to the field of underwater target tracking. For underwater targets, because their motion patterns are not fixed, they often maneuver, so it is difficult to keep the parameters of the tracking estimation filter consistent with the actual situation, which will lead to a large increase in tracking error when the target maneuvers, and the tracking effect will deteriorate. It may cause the tracking error to diverge and thus lose the target. In view of this problem, the method proposed by the present invention uses the support vector machine to judge the current movement pattern of the target, and corrects the conditional model probability of each movement model in the interactive multi-model algorithm according to the classification result, so that the interactive multi-model algorithm can Quickly respond to the target's movement mode switching, reducing the tracking error and ensuring the convergence of the tracking. Therefore, the present invention has important practical significance for solving the problem of underwater maneuvering target tracking.

Description

Technical field [0001] The present invention relates to the field of target tracking, and more particularly to an interactive multi-model-based support vector machine assisted water machine target tracking method. Background technique [0002] The ocean contains rich natural resources, one of the important concerns of human future development, and is also an important stage of national integrity. Underwater target tracking technology in military fields, such as detection of threat targets such as torpedoes, submarines, frog people, and submerged missiles; civilian fields such as search and rescue, salvage, unmanned navigation, etc., scientific research, such as marine life tracking , Environmental monitoring, etc., all have important applications, have broad development prospects. [0003] The tracking of underwater goals is mainly done by ship drag and underwaterwireless sensor networks (UWSNs). Significantly obtain an analog signal in a proactive or passive manner, extract the ...

AI Technical Summary

Problems solved by technology

However, when the actual motion pattern of the model changes, the tracking effect will inevitably deteriorate during the transition phase

Since the interactive multi-model method believes that the switching between models obeys the first-order Markov process, it is difficult to prove that the motion of the tracked target in the actual scene strictly obeys this process, so the historical information of measurement and estimation is difficult to be fully utilized

Images