Research On Ground Multi-source Data Fusion Recognition Of Command System

The command system is an important support for the implementation of command and control on modern platforms.Multi-source data fusion identification is to provide data support for command decision-making.Multi-source data fusion technology has matured with the continuous development of science and technology,and is widely used in many fields.Different systems have different backgrounds,have different structures and requirements,and use different data fusion models and technologies.This article first briefly introduces the principles,functional models and key technologies of multi-source data fusion recognition,and then studies the following key technologies in multi-source data fusion recognition according to the system requirements and data characteristics of this article.The main research contents are as follows:(1)System plan analysis.The system studied in this paper needs to perform data fusion processing on the measurement information transmitted by each data collection and analysis platform to obtain target identity information and travel path information.Analyze several common data fusion structures,choose a centralized fusion structure design system plan,and apply data correlation technology and target recognition technology to complete system functional requirements.(2)Research on data association method based on speed direction.The role of data association is to determine the affiliation between measurement information and target,to perform data preprocessing on the message data,that is,to perform spatial alignment and time registration,so that the data is under the same reference standard,and provide data support for data shutdown calculation.Aiming at the problem that multi-moving targets are prone to dense paths and cross paths,a nearest neighbor data association method based on velocity direction is proposed to achieve accurate data association and lay a foundation for target recognition.(3)Target recognition method based on evidence support level.Target recognition is to analyze and calculate all the information of a certain target to determine the target identity.Aiming at the problem that conflict evidence in traditional D-S evidence theory affects fusion results,a target recognition method based on evidence support level is proposed to reduce the impact of evidence conflict on fusion results and improve the accuracy of fusion results.(4)System software design.According to system requirements,five functional modules of communication,database support,data association,and target recognition are designed to complete multiple functions from receiving multiple data collection and analysis platform messages,measuring information and target association,and target identity judgment.Design a human-computer interaction interface to observe data association and target recognition results.(5)Experimental analysis.The system experiment platform is used for data association and target recognition processing on the ground multi-source data information to analyze the experiment data.Through comparative experiments,the feasibility of this method is verified.