Distribution estimation method

Abstract

The present invention discloses a distribution estimation method. The method comprises the steps of: Step 1, when a current object is subjected to imaging each time, extracting prospect information provided by an object discovery satellite by employing an object identification satellite; Step 2, deleting objects with low observation values in the prospect information according to three threshold parameters in object filtering knowledge; Step 3, calling a plurality of heuristic algorithms concentrated by a scheduling algorithm to generate a plurality of local observation schemes, wherein the step can be subjected to parallel calculation, and the local observation schemes are observation schemes generated by the satellite according to object information in a prospect time window in a currentstate; Step 4, performing comprehensive assessment of each local observation scheme according to an assessment parameter combination correlation assessment method in scheme assessment knowledge; andStep 5, selecting a local observation scheme with the highest score and locking a first object in the schemes as the next observation object.

Description

technical field [0001] The invention relates to the field of satellite technology, in particular to a distribution estimation method. Background technique [0002] The current scheduling method of agile satellites is to set up a double star cluster consisting of a low-resolution target discovery star (full name "target discovery satellite") and an agile high-resolution target recognition star (full name "target recognition satellite"), The target finds that Xingfei is at the front of the star cluster, and uses a low-resolution camera to perform push-broom imaging of large areas on both sides of the line below the star, and can extract relevant target information in the imaging area in real time. The target identification star receives the target information extracted by the target discovery star in real time, and obtains the position information, imaging duration, and observation income of the target 100s in advance. When the target recognition star is imaging a certain tar...

AI Technical Summary

Problems solved by technology

[0003] However, since the target recognition star can only obtain the local information of the target distribution (only the target information in the look-ahead time window, not the task distribution information of the entire scene) each time it makes a decision, it is easy to determine the next observation target every time. Cause "short-sighted" problems

Especially when the scene period becomes longer and the satellite’s fixed storage and power constraints are tight constraints, these resource constraints may not be able to support the target recognition satellite to image too many targets in one orbit (high-resolution images occupy a large storage space )

This kind of "short-sighted" decision-making will lead to the premature consumption of on-board resources such as fixed storage by the target identification star, which may cause the satellite to give up some high-yield targets due to insufficient fixed storage and other resources in the later stage, thereby reducing the Global Observation Benefits of the Cluster

Images