Particle Filter Track-before-detect Algoirthm Applied To Weak Targets Detection

The TT&C system of deep space exploration functions as telemetry, tracing andcommand for the faraway detector in space, which keeps a reliable communication anddata transmission with the object so as to ensure the orbit accuracy and acquire thescientic information. For the actual requirements of some project, the detection andtracking theoretical algorithm, which is based on particle filter tracking beforedectection(PF-TBD) algorithm to detect the weak signal in TT&C system, is presentedin the thesis.TBD is an important method for the detection and tracking of weak targets.Different from classical track-after-detect(TAD) processing, TBD directly usesunthresholded or low thresholded measurements of sensors for adequate information. Itgains the increasing of signal to noise(SNR) by temporal cumulation of measurementsand realizes the detection and tracking(track extracting) of weak targets simultaneously.Since TBD usually behaves as a hard nonlinear problem, PF-TBD algorithm isinvestigated. This thesis focuses on research to solve non-linear, non-Gaussian problemof particle filter.Around the particle filter in the two most critical importance are sampling densityfunction and weight problems, the Kalman Particle Filter and Extended Kalman Filteralgorithm are analysised, and then an improved particle filter is given, then re-samplingMarkov chain, combined with improved particle filter algorithm, is analysised in thedissertation. To solve the practical application of particle filtering computationallyintensive problem, an improved particle filter is designed and the feasibility of hardwareFPGA implementation is analyzed. Observation model for the design of the sensorpre-test PF-TBD method and a kind of analysis simulation are given.