Differential evolution method facing agile satellite multi-object task planning

Abstract

The present invention discloses a differential evolution method facing agile satellite multi-object task planning. The method comprises: converting a solution set space formed by decision variable to a population; performing initialization of the population and the algorithm parameters of the population; generating the donation vectors, trial vectors and filial generation according to the algorithm parameters of the population; adding the filial generation into the population to obtain a variation population, and obtaining the fitness of an individual decoding; determining whether the current iterative times i is smaller than the sum of the iterations or not, if the current iterative times i is smaller than the sum of the iterations, performing selecting of the individuals in the variation population, generating a new population, and updating the algorithm parameters of the corresponding population; or else, rejecting the controlled solution in an elite solution set, adding the non-control solution in the variation population into solution controlled by the individuals in solution set without elite to update the elite solution set; and sorting the individuals in the elite solution set, and outputting the individuals with the assigned number according to the sequence. The differential evolution method facing agile satellite multi-object task planning can integrate one-body analysis of the multi-object optimization characteristics.

Description

technical field [0001] The invention relates to a differential evolution method for agile satellite multi-objective task planning. Background technique [0002] Agile satellite multi-objective programming problems have obvious time-dependent characteristics, and analyzing the research status of time-dependent scheduling methods is of great help to the research of agile satellite multi-objective programming. The current research on time-dependent scheduling methods mainly focuses on three aspects: time-dependent processing time, time-dependent revenue, and time-dependent switching time. Gupta and Browne first proposed time-dependent processing time. Subsequent scholars studied the extended form of the problem. The objective function of the research is mainly divided into minimizing the sum of the completion time and minimizing the processing cycle, and the dependence of processing time on the processing start time. The relationship is mainly linear, and scholars such as Huan...

AI Technical Summary

Problems solved by technology

When facing the time-dependent conversion time problem, Pralet uses a time-dependent simple time network to deal with it, corresponding to the use of constraint propagation in agile satellites to adjust the start time of the next imaging task, but does not consider the optimization of the problem Model and Algorithm

Chen Cheng studied the workpiece processing problem with time-dependent conversion time and time-dependent income, designed a hybrid difference algorithm, and applied it to agile satellite imaging task planning, which has a good solution effect, but the defect is that the considered target is only Single target planning for point target imaging cannot be applied to multi-target situations

Images