Mutiple Sensor Maneuvering Targets Tracking Algorithm In Air Tests

Multiple maneuvering targets tracking problem is a research focus of the field oftarget tracking at present, and it has extensive application perspective in military andcivil industries, such as air warning, battlefield surveillance, air traffic control, andnavigation systems, etc. With the rapid development of science and technology, moderntrack environment has undergone significant changes. Multi-targets tracking and dataassociation technology faced strong challenges. This paper is concerned with twoaspects of multiple maneuvering targets tracking problem through the weapons andequipment pre-research projects, those are maneuvering target tracking and dataassociation. The main work and achievements are as follows:Firstly, an overview of the basic principles of maneuvering target tracking ispresented. The basic elements of maneuvering target tracking problem are summed up,including the target motion models and the commonly used filters for maneuveringtarget tracking algorithms. The paper also analyzed the characteristics of the variousfiltering algorithms.Secondly, to implement the problem of multiple non-maneuvering targets dataassociation and track maintenance in cluttered environment, the traditional associatingmethods such as joint probability data association (JPDA) and new associating methodssuch as graph-based and biology-based association methods are studied. And two kindsof data association algorithms including the JPDA and m-best MHT are implemented.Taking into account the high computational complexity of the m-best MHT algorithm,two improved m-best MHT algorithms are proposed by reducing the dimension of therow vectors and column vectors of the cluster matrix. Respectively, the proposedalgorithms perform better in the environment which has large number of targets andmeasurements. Monte Carlo simulation results show that the algorithm can shorten thetime with maintaining the filtering accuracy and association accuracy.At last, to implement the problem of multiple maneuvering targets data associationand track maintenance in cluttered environment, two kinds of algorithms includinginteractive multiple model joint probabilistic data association (IMMJPDA) andinteractive multiple model multiple hypothesis(IMMMHT) have been researched andimplemented. And a improved IMMMHT algorithm are proposed by grouping themeasurements, and greatly eased the problem of enormous calculation caused by splitting the high-dimensional cluster matrix. By two and three maneuvering targettracking simulation results show the effectiveness of the above algorithm.