关键词: 知识工程+;知识重用+;一般模糊极小极大神经网络+;固体推进剂火箭发动机;结构模式识别+

Abstract: To discuss the application of knowledge reuse in solid rocket motor(SRM) design,general fuzzy min-max(GFMM) neural network was used for structure selection of SRM design and knowledge-based engineering and fuzzy sets theory were studied as well as automatism and quantification of SRM structure selection were realized.Firstly,the model based on knowledge reuse of SRM structure selection was presented.Secondly,GFMM neural network for SRM structure selection was created.Finally,structure selection was finished in a SRM system design by the GFMM neural network.Many feasible results of structure selection were gained,which included the actual result.This technology can improve the degree of automatism and quantification of SRM structure selection,all of different feasible results are provided for SRM designer.

Key words: Knowledge-based engineering+;Knowledge reuse+;General fuzzy min-max neural network+;Solid propellant rocket engine;Structure pattern recognition+