论文部分内容阅读
在模具标准制修订工作中,是否采用国际标准或国外先进标准、预测标准对国情的适用性是重要环节。通过对我国现行模具标准的大量分析,确定标准评价因素,构建标准评价体系,运用层次分析法(AHP)计算各因素的重要程度值,并做量化分析和评分,以获得线性神经网络模型样本,进而对样本进行训练、验证,最终获得模具标准评价模型。该模型充分吸收了专家的知识和经验,降低了评价的人为因素。结果表明,基于AHP和线性神经网络的模具标准评价方法计算的最大相对误差为1.6%,该评价方法正确可行。
In the mold standard revision work, whether the use of international standards or advanced foreign standards, prediction of the applicability of national standards is an important part. By analyzing a large number of current mold standards in our country, the standard evaluation factors are determined, the standard evaluation system is established, the importance of each factor is calculated by AHP, and the quantitative analysis and scoring are done to obtain the samples of linear neural network model, And then the sample training, verification, and finally get the mold standard evaluation model. The model fully absorbed the expert knowledge and experience, reducing the evaluation of human factors. The results show that the maximum relative error of calculation based on AHP and linear neural network is 1.6%, which is correct and feasible.