Time domain evidence fusion method based on adaptive processing strategy

Abstract

The invention discloses a time domain evidence fusion method based on an adaptive processing strategy, and the method comprises the following steps: firstly, converting information collected by a plurality of sensors into evidence information; secondly, calculating the correlation of the evidence information at each historical moment, and determining the weight factor of the historical evidence information; performing weighted average on the historical evidence information to obtain a predicted value at the current moment, and judging whether distortion occurs at the current moment according to the predicted value and a true value; obtaining the real-time reliability of the evidence at the adjacent moments through a credibility attenuation model, and carrying out the discount of the time domain evidence according to the real-time reliability; and finally, fusing the time domain evidences by using a Dempster combination rule. According to the method, the situation that the evidence at the current moment is likely to be distorted is considered, meanwhile, the distortion situation is researched, the evidence obtained at the current moment is compared with the historically accumulated evidence, the most appropriate processing strategy is selected, and the method has important theoretical significance and application value.

Description

technical field [0001] The invention relates to the field of target recognition, in particular to a time-domain evidence fusion method based on an adaptive processing strategy. Background technique [0002] In the air defense and anti-missile combat system, target identification is an essential and important link, and the accuracy of its results directly affects the accuracy of situational assessment and threat estimation, as well as the commander's strategic deployment. With the rapid development of weapon technology, the confrontation of modern air defense and anti-missile battlefields has been continuously enhanced. The complex electromagnetic environment and natural environment often lead to great uncertainty in the information on the battlefield, and the target information obtained through sensors may be distorted. The phenomenon. Therefore, it is not only necessary to perform spatial fusion of the information provided by each sensor in a single measurement period, but...

AI Technical Summary

Problems solved by technology

Due to the obvious sequential, dynamic and real-time nature of time domain evidence fusion, the method of air domain evidence fusion cannot be used directly to fuse time domain evidence, which will lead to information loss

[0004] The reliability decay model in literature [1] fully considers the influence of time factors on the fusion of time-domain evidence, and proposes a relative reliability factor based on the reliability decay model, but there are deficiencies in the evaluation of time-domain evidence

[0005] Literature [2] considers the real-time reliability evaluation based on the reliability decay model when performing time-domain evidence fusion, but does not consider the relative reliability evaluation of the evidence, which is easy to cause the loss of evidence information

[0006] Literature [3] combines the intuitionistic fuzzy set with the reliability decay model in the fusion of time domain information. Since the reliability decay model believes that the evidence at the latest moment has the greatest reliability in the reliability setting, the reliability of the latest evidence is If the degree factor is set to 1, the reliability factor of the evidence at the previous moment is quantified by the credibility decay model, but if the evidence information obtained by the sensor at the latest moment is distorted, it is still given the maximum weight at this time, which will bring greater error

[0007] The above methods all have the problem of easy loss of evidence information when dealing with time domain information.

Images