Based On Variable Structure Multiple Model Algorithm For Target Tracking System

In many fields, like in the field of military air-defense systems, battlefield surveillance, civil aviation air control, vehicle navigation and so on, target tracking technology has been widely used today. Along with the accurate target tracking demand increasing, as well as a variety of high speed, high maneuvering aircraft unceasing emergence, the target tracking technology research becomes a hot spot of scientific research now.This thesis thoroughly studies the way to design and realize the target tracking system based on the variable structure multiple model algorithm. At first, this thesis introduces the principle of target tracking. In the next part, this thesis researches the method to select filter algorithm as well as the solution of the filter divergence phenomenon which appears in practices frequently, and according to the actual requirement of tracking system, a mix filter algorithm is proposed in this thesis. The simulation result shows that, its track effect has the very big enhancement compared to the standard Kalman filtering. Then this thesis thoroughly researches the model group switch (MGS) algorithms. And a thorough research is done on the core part of the MGS algorithm. A simple helicopter tracking system is designed completely with this algorithm. Finally, the software is designed completely of this tracking system which base on the MGS algorithm, and uses some common maneuvering forms to verify the performance of the tracking system.The variable structure multiple model algorithm as a kind of new development direction of the target tracking algorithm, overcomes all sorts of flaws of the formerly fixed structure multiple model algorithm, accords with the demand of the target tracking algorithm with higher performance. In the end of this thesis, some questions about the current statistical model, filter algorithm and MGS algorithm are proposed as the continuing study directions.