Method for distributing sensors in collaborative detection

Abstract

The invention provides a method for distributing sensors in collaborative detection. Based on the optimal Kalman distributed filtering, a new information entropy incremental computation method and a global optimal distribution scheme are put forward, only the optimal solution of a distribution equation formed by new information entropy increments, serving as parameters, of all targets to all single sensors under the constraint condition is calculated, and sensors attaching to the same target under the optimal solution form a sensor group of the target. The new information entropy incremental computation method is put forward, so that the problem that a covariance matrix of a target under distributed fusion filterers is not just linearly summed through a sensor group formed by respective covariance matrixes of multiple single sensor filters and multiple single sensors is solved.

Description

technical field [0001] It involves multi-target multi-sensor management technology. Background technique [0002] In the tracking system, it is of great research significance to rationally and effectively allocate and manage multiple targets distributed on different platforms, different types, and different performances under many constraints. In practical applications, limited sensor resources require reasonable sensor management to meet the requirements for multiple targets and scanning spaces, obtain the optimal detection value of each feature, and then evaluate the optimal feature value. The core problem of sensor management is to allocate corresponding sensors for the target and set the working mode and parameters of the sensors according to certain optimal criteria. The rapid development of sensor technology, signal detection and processing technology has greatly promoted the progress of multi-sensor data fusion theory, and its technology has gradually been widely use...

AI Technical Summary

Problems solved by technology

[0004] Kalandros and Pao considered the sensor allocation problem from the direction of covariance control, and calculated the expected covariance matrix as the cost coefficient to control the allocation of sensors. This method of sensor dynamic allocation is more targeted and flexible. All possible groups, and then predict all targets for all possible groups. When the number of sensors and targets increases sharply, combination expansion inevitably occurs, which consumes a large amount of calculation

Images