摘要：The continuous developing features in the design of mechanical product and based on 3D entity model is aimed at, and the extension of the 4-dimensional model with the process of designing, the knowledge described model on the level of semantic understanding and summarizing the designing process and the way of discovering knowledge from multi-information model are studied. On the basis of designing the broad sensed collaborative system, through discussion of the relationship between the implicit knowledge of the users and the designing knowledge as well as commanding all the designing links, taking advantage of the way of concluding and deducting in the concept of the designers, the synthetic knowledge unit formed in the dynamic process from the conception design t9 the last design is schemed out, and the knowledge discovered principle in the dynamic designing process of the mechanical products and the key technology in its implementation under the milieu of network is brought forward.

摘要：针对北斗某星辐射剂量探测数据缺失问题,提出了一种基于线性样条和CNN-LSTM神经网络模型的处理方法。在对数据特性分析的基础上,将原始数据分解为线性趋势项和季节波动项。对于线性趋势项,采用基于线性样条的缺失值处理方法;对于季节波动项,根据其时空变化特性,设计CNN和LSTM组合神经网络结构,完成季节波动项的缺失值处理。实验表明,相比于线性插值法和傅里叶变换插值方法,本文所提方法的插补值与真实值偏差更小,相关性更高。平均相对误差达到0.008,相关系数达到0.855。同时横向对比了本文所提组合神经网络模型和单一的LSTM和CNN网络模型的插补结果,同样本文方法表现出更好的一致性。研究结果表明,本文方法能够较好解决北斗数据连续缺失的问题,为后续基于北斗数据开展科学研究和业务应用奠定基础。