The General Optimization Design Of Small Satellites Based On Cloud Model

As an important part of the general design phase of small satellites, theoptimization design of small satellites runs throughout the whole general design process.The optimal results have a profound impact on the prelinary design and detailed designof small satellites, which directly determines the quality and efficiency of the satellitedesign. In order to meet the design needs of small satellites, establishing rationaloptimization models and designing efficient optimization algorithms have an importanteffect on improving the efficiency of small satellites design, shortening the developmentcycle and reducing the development cost. Supported by the Eleventh Five-years DefensePre-research topics and National High Technology Research and Development Program863, the optimization design of small satellites are studied deeply in this thesis whichmainly focuses on the problems about how to set up the optimization models meetingthe requirements of small satellites design and how to design efficient optimizationalgorithms. The main contents of this thesis are consisted of the following parts:As the optimization design of small satellites is a mixed variables optimizationproblem containing both continuous varialbes and discrete variables and its solutionspace is often non-convex and not connected. With the observation missionrequirements of sun-synchronous earth observation small satellites, a optimizationmodel of small satellites quality is established. Combining evolutionary algorithms withCloud Model(CM), which is wide used in In the field of artificial intelligence, a newoptimization method named Evolutionary Algorithm Based on Cloud Model(EABCM)is proposed. The effectiveness of EABCM method is verified by the optimizationproblem of the small satellites quality.In view of the requirements about the optimization design cost of small satellites,on the basis of analyzing and comparing. with the common cost models of smallsatellite, according to the requirements of earth observation mission, a new optimizationmethod named Particle Swarm Optimization Based on Cloud Model(PSOBCM) isproposed, which solve the defects for particle swarm optimization about convergingslowly and being easy to fall into local extreme point. By numerical simulation theanalysis results in theory about the accuracy of the cost models and the effectiveness ofthe proposed PSOBCM method areverified.As for the higher requirements about reliability design of small satellites, firstly theconcepts and design steps of reliability design of small satellitesas well as the estimationmethod about the reliability are analyzed; secondly according to the needs of missiondesign, a multiobjective optimization model about the reliability and design cost of thesmall satellites is set up. To solve this multiobjective optimization problem, using thesorting idea of the multiobjective genetic algorithm, a new optimization method named Multi-Objective Evolution Algorithm Based Cloud Model(MOEABCM) is designed. Atlast the effectiveness of the MOEABCM method is verified by simulation results.Given the requirements about the rapid response space mission, firstly the generaldesign ideas and methods of the upper-stage spacecraft is analyzed, and design steps ofthe subsystem for the upper-stage spacecraft is also shown; secondly according to thedesign objective of the upper-stage spacecraft, the quality optimization model, costoptimization model, cost and reliability multiobjective optimization model of theupper-stage spacecraft are respectively founded; at last using evolutionary algorithmbased on cloud model, particle swarm optimization based on cloud model andmulti-objective evolution algorithm based cloud model respectively, three generaloptimization design models are optimized. The effectiveness of the optimizationmethods based on cloud model and the rationality of the upper-stage spacecraft designare also verified.