Dynamic multi-objective optimization method based on segmentation multi-directional prediction strategy

Abstract

The invention discloses a dynamic multi-objective optimization method based on a segmentation multi-directional prediction strategy, which can better predict the change of an optimal leading edge, ensure the prediction of the distribution change of the concavity and convexity of the optimal leading edge and effectively compensate the distance error of direction prediction in allusion to the problem of insufficient prediction error compensation of a linear model. The method comprises the following steps: (1) describing a dynamic multi-target problem; and (2) carrying out segmentation multidirectional prediction. The beneficial effects of the invention are that: t the method comprises the following steps: firstly, segmenting a searched optimal leading edge; dividing the population accordingto the segmentation result of the optimal leading edge, carrying out linear prediction on each part of the population, referring to the description of a cloud model for uncertain time, compensating prediction errors by using entropy and hyper-entropy, and adopting a deflection angle enhanced search strategy to ensure the diversity of the population in the prediction process aiming at possible sudden nonlinear changes in the dynamic problem.

Description

technical field [0001] The invention belongs to the technical field of intelligent computing and multi-objective optimization, and in particular relates to a dynamic multi-objective optimization method based on segmentation and multi-directional prediction strategies. Background technique [0002] In the process of space and space high-speed target detection and tracking, due to the fast target speed, strong maneuverability, and large flight area span, it is difficult for a single sensor to detect and track effectively and continuously. Therefore, it is particularly important to use multi-source heterogeneous sensor scheduling technology to reasonably arrange detection and tracking tasks and timing. important. The dispatched resources mainly include space-based infrared sensors and ground-based radar sensors. The types of sensors involved include high-orbit infrared detection satellites, low-orbit infrared satellites, ground-based P-band radars, and ground-based X-band radar...

AI Technical Summary

Problems solved by technology

[0005] (1) The task has no time to wait, and if the target is not observed within a limited time, there is no opportunity to observe

[0006] (2) One-to-many and many-to-one resources matching exist at the same time, and the inconsistency of matching rules makes encoding and searching more difficult

[0007] (3) The observation time of the sensor is fragmented, and it is easy to generate fragmented time that does not meet the constraints during the search process, making the search more difficult

However, it is worth noting that although the linear model is simple, there are always prediction errors in solving dynamic problems of increasing and decreasing or more complex changes. Therefore, the content disclosed in the above-mentioned prior art uses the method of maintaining diversity to make up for the prediction error

Images