Multi-objective Bernoulli Distributed Fusion Method Based on Weighted Negative First Order rd Sum

Abstract

The invention discloses a multi-target Bernoulli distributed fusion method based on a weighted negative first-order RD (Renyi Divergence) sum, belonging to the field of radar signal processing. By transmitting the posterior multi-target Bernoulli parameters of the same frame of multiple sensors to the fusion center, the posterior multi-target Bernoulli parameters of each sensor are fused in the fusion center. The fusion method is to combine the multi-target Bernoulli parameters Decompose into multiple Bernoulli components, find the optimal matching scheme of Bernoulli parameters of different sensors by constructing a cost function based on the weighted negative 1st-order RD sum, and then use the CI fusion criterion to fuse multiple matching Bernoulli in parallel Components, approximately realize the fusion of multi-target Bernoulli parameters of different sensors, and then select the most likely multiple targets from the fusion results and return the fused results to each sensor, so that it has high tracking accuracy, low false alarm rate, Strong robustness, high reliability, low communication cost, and strong anti-interference ability.

Description

technical field [0001] The invention belongs to the field of radar signal processing, and particularly relates to a multi-target tracking technology and a multi-target Bernoulli distributed fusion technology based on a negative first-order weighted RD sum. Background technique [0002] Due to the stealth design characteristics of modern targets (or electromagnetic interference environment), the single-view observation of single-node radar systems cannot meet the needs of effective target detection, prompting domestic and foreign engineers and researchers to use multi-node radar networking systems to solve stealth target detection. difficult question. The multi-node (sensor) radar system obtains the measurement information of the target at different node positions through multi-view observation. In order to effectively use the information provided by each sensor to track the target, it is necessary to design a high-reliability, low-computation, and easy-to-implement Multi-se...

AI Technical Summary

Problems solved by technology

However, after derivation, we found that the mixed exponential density after the fusion of two multi-objective Bernoulli distributions does not obey the multi-objective Bernoulli distribution, so the distributed fusion of multi-objective Bernoulli filters cannot be realized directly by using the CI fusion criterion

Images