Multiple Hypothesis Composite Tracking Algorithm Research

The trend of the development of warfare in the future is a kind of battle mode centered on the network and the traditional flat centric warfare is difficult to satisfy the need of development of advanced weapon. As a result, many countries in the world transit the focus of research on Network Centric Warfare (NCW) step by step. The most attractive characteristic of NCW is that it actualizes the Cooperative Engagement Capability (CEC) by sharing the warfare development situation in highly real-time. The composite tracking algorithm discussed in this paper is the first step to accomplish the CEC. The capability of discovering the hidden targets such as cruise missile skating over the sea can be strengthened by the multi-sensor network. The radar blind area can be mostly decreased and the composite situation picture can be built in order to make the timely and efficient response.People used to research on the problem of track fusion and seldom focus on the composite tracking. Because the new concept is proposed under the new system of NCW. In recent years, the most of the published paper about NCW in foreign countries barely put forward the concept and principle of composite tracking algorithm and hardly discuss the specific algorithm of composite tracking. The research on the composite tracking in China is on the beginning stage and papers on the composite tracking in the nation are mainly the translation of the foreign document and the complete system of theory has not been built. The main research on multisensor composite tracking is focus on non-maneuvering target. With the background of NCW, this paper mainly concentrates on the maneuvering target multiple hypothesis composite tracking algorithm in order to solve the problem of the maneuvering target composite tracking. This is a new field of research and no deep research has been started.In this paper, some new tracking algorithms have been introduced including the unscented Kalman filter used in the research of this paper and the model of motion and observation of non-maneuvering and maneuvering targets is given. Then the key algorithm of this paper -- multiple hypothesis composite tracking algorithm is proposed and the principle and framework are introduced and three type of tracking gate is discussed. The core idea of multiple hypotheses composite tracking algorithm is based on the concept of multiple hypothesis tracking theory and through the extrapolation of multiple hypothesis tracks building the area of space association of composite tracking and make a decision whether the tracks are ready to composite by judging the relationship between this area and the associated beginning track. The other key problem of multiple hypothesis composite tracking algorithm is the composite tracking association criterion. In this paper, based on time, space and track attribute, a new association criterion is proposed. This criterion not only considers the association in the space of detection but also includes the association of time and attribute. The research indicates that this new criterion is favorable for solving the problem of composite tracking, especially the composite tracking of maneuvering targets. The result of simulation indicates that the multiple hypothesis composite tracking algorithm can composite non-maneuvering and maneuvering targets preferably which satisfy the criterion and solve the problem of long range and super long range target composite tracking.