Research On Radar Object Classification Algorithm Based On Kinetic Information And RCS Features

Radar target classification and recognition has important applications in the fields of military and defense, such as air and missile defense, maritime defense. Radar, as an important tool of acquiring war information, is monitoring the entire battlefield all the time. Its target classified recognition can be used to the situation assessment and decision analysis in the battlefield, which can improve our combat effectiveness and battlefield survivability.In our country the great mass of active service radar is narrowband low-resolution, and its ability of getting target information is relatively limited, which is usually characterized by its target position, velocity, acceleration, RCS (Radar Cross Section) for target classification. In this paper, the research of target classification is mostly based on kinetic information and RCS features, including the following three parts:1. This paper proposes a joint track classification algorithm for the inaccurate problem of unknown maneuvering target model in the radar target tracking process. Based on the feedback of information in the tracking process, the multiple target maneuvering model pre-established can be classified by target probability and mobility. And these classification results will assist correct target maneuvering model, which can be classified by the mobility and improve the tracking accuracy simultaneously.2. This paper presents a moving target classification algorithm based on fuzzy logic for the insufficient mobility of some target in the smooth phase of flight and the possible inaccurate classification in the joint tracking of the target classification algorithm. This classification algorithm uses Fuzzy mathematical knowledge, blurs the target kinetic information and generates the fuzzy relationship through the expertise, after a certain time puts the normalized target height and speed values into the established fuzzy logic inference system and finally estimates classification probability of all kinds of targets.3. This paper proposes a target classification algorithm based on kinetic information and RCS feature classification fusion for its limitation in solely depending on the kinetic information in the target classification process. Algorithm is going to classify based on the target kinetic and RCS feature, and then use the DS evidence theory fusion to classify at the decision-making level. This classification algorithm makes full use of the radar echo information, and fuses the historical results as well in the process of fusion, which effectively improves the Algorithm's convergence speed and robustness.