Multi-sensor Distributed Data Fusion Method Based on Chernoff Fusion Criterion

Abstract

The invention discloses a multi-sensor distributed data fusion method based on Chernoff fusion criterion. It first performs particle filtering on each sensor to obtain local estimation results, uses the maximum expectation algorithm to approximate it as a Gaussian mixture distribution, interacts Gaussian mixture parameters between multiple sensors, and uses the Chernoff fusion method under the first-order approximation model to obtain multi-sensor The preliminary fusion results of the results, the results of the importance of sampling, the construction of multivariate variables with constraints of the optimization function, using the particle swarm optimization algorithm to solve, calculate the index value of the multi-sensor under the Chernoff fusion criterion, and finally use the particle sample Fusion with Chernoff optimal index weights for Chernoff fusion to calculate the estimated state of the target. The present invention theoretically and approximately optimally solves the problem that it is difficult for any number of multi-sensors to obtain ideal distributed fusion results under the condition that local estimation results are correlated.

Description

technical field [0001] The invention belongs to the technical field of multi-sensor data fusion, and in particular relates to a multi-sensor distributed data fusion method based on Chernoff fusion criterion. Background technique [0002] With the increasing complexity of the modern battlefield environment, the urgent need for stealth and anti-stealth, confrontation and anti-confrontation, and the emergence of problems such as strong maneuverability, high clutter, low detection rate and high false alarm rate, using multi-sensor data fusion to obtain more Comprehensive, accurate and reliable environmental situation information has attracted more and more people's attention. Among them, distributed data fusion has been greatly developed due to its many advantages such as low communication volume, strong scalability, and good robustness, and has been widely used in many fields such as area monitoring, target tracking, and target positioning. [0003] For distributed data fusion...

AI Technical Summary

Problems solved by technology

The prior art proposes a Chernoff fusion method under the first-order approximation model, which can finally obtain a convergent fusion result by performing exponentially weighted fusion on the estimated results of different sensors; however, this method introduces Some unreasonable assumptions and approximation processes lead to large information loss in fusion results and low fusion accuracy; the existing technology also proposes a Chernoff fusion method based on importance sampling, which is also based on exponential weighting Therefore, it can effectively solve the problem of correlation of local estimation results; in addition, this method adopts the strategy of importance sampling without any approximation process, so there is almost no loss of accuracy; however, this method is only applicable to Fusion between two sensors, and the iterative fusion of two sensors for multi-sensor fusion will lead to the index weight of Chernoff fusion as a suboptimal solution, and the accuracy of fusion results will be greatly reduced

Therefore, this method is not suitable for distributed fusion of any number of sensor network systems

Images