Research On Analysis Of Maritime Moving Target For Satellite Resource Scheduling

Research on analysis of maritime moving target for satellite resource scheduling is of great applicational value in some domains, such as maritime rescues, anti-smuggling and situation assessment. The problem of research on analysis of maritime moving target for satellite resource scheduling is complex due to the complexity of observing environment, the uncertainty of maritime moving target and facility faultiness. However, it has many open issues in theory and practicality to be tackled.This paper focuses on the scheduling problem of maritime moving target search by satellites resource, which includes problem formulation and analysis, planning model establishing, target state estimating, sensor scheduling, and so on. The main work and of this dissertation can be concluded as the following four parts:(1) On the base of analyzing maritime moving target search for satellite resource scheduling, we indicate defects of existing work and the characteristic of the problem. A solving framework is proposed which contains target state estimating and sensor scheduling.(2) In estimating of target state we introduce Unscented Particle Filter (UPF) theory. In the single target case,we combine Interacting Multiple Model (IMM) in the mobile target tracking and propose a IMM-UPF algorithm for target state estimating. In the multi-target case, we adopt Joint Multitarget Probability Density (JMPD) to denote the target state and propose a JMPD-UPF algorithm for target state estimating.(3) To solve the sensor scheduling problem, we propose a scheduling algorithm based on Rényi Divergence according to the information theory. We define the entropy as variation of probability density in the area. In the scheduling process, our algorithm selects the area with maximum entropy to get the maximum information gain and detect the target more possiblly.(4) Finally, we propose algorithm for single target and multi-target according to the research result above, and then design and implement a prototypal system. Some experiments are conducted to validate the proposed method and demonstrate its practicability.