基于检测特征的车身覆盖件检具概念设计

Concept Design of Checking Fixtures for Auto-body Parts Based on Measuring Features

首次提出了车身覆盖件检测特征的概念,在讨论这个概念的基础上,采用人工神经网络进行车身覆盖件检具概念设计,以检测特征的7个分量作为神经网络的输入,以检具类型分量作为输出,对构成的神经网络用生产中的实例作为样本进行训练,达到误差平方和小于0.01的目标,得到检具概念设计神经网络模型,并以具体零件检具概念设计为例验证了该方法的有效可行。

Checking fixtures for auto-body parts are designed with neural network based on the discussed measuring features from auto-body parts. The input of neural network model is composed of 7 sets of measuring features AND the output is composed of 4 types of checking fixtures. In order to obtain error （between output of sample and model） less than 0.01, the neural network model is trained by some real example dates, then an example is used to validate the model for selecting types of checking fixtures. The results shows that the model can select optimal type of checking fixture for auto-body part.