中文摘要：

针对低碳设计多维度实例检索建模及其降维计算问题,提出了一种多维关联函数建模及其检索应用方法。首先,通过变斜率原理改进侧距,构建了基于改进侧距的多维侧距模型。然后,针对侧距难以准确描述设计参数与最优值之间距离的问题,建立了多维关联函数及其三层次的低碳实例检索方法。最后以螺杆空压机为例进行了验证,结果表明了改进侧距的正确性和有效性、多维关联函数降维计算的快速性与有效性,以及低碳实例检索的合理性与实例间的差异性。

英文摘要：

To solve problems of multidimensional case retrieval modeling and its dimensionality reduction calculation for low carbon design,a modeling and application method of multi-dimensional correlation function was proposed.Firstly,an improved side distance method was developed through the principles of variable slope,which provided the basis for developing a multi-dimensional side distance model.Secondly,because side distance formula could not give out the accurate distance values between design parameters and its optimal,a multi-dimensional correlation function which had good dimension reduction effect was established.Then a retrieval method was given based on the multi-dimensional correlation function.Finally,the proposed method and algorithms were verified in the design of a screw air compressor.The evaluation results show that the improved side distance is correct and effective,the dimension reduction method for the multi-dimensional correlation function is fast and viable,and the otherness among these cases is very obvious.