论文部分内容阅读
一、从电池寿命预报说起宇航用的电池对寿命要求特别高,是不言自明的事。但由于密封Ni-Cd电池(一种钮扣电池)在应力下使用时寿命波动很大,同类产品间可相差一倍之多。所以需要通过抽样实行加速破坏试验来检验质量。加速试验分八个单项因子测定。如何从这八套数据来关联或预测电池寿命呢?美国普图大学的布耶(Buyer)等采用电子计算机模式识别(Pattern Recognition)技术,成功地解决了这一课题。模式识别又名“图形识别”。原来“图形识别”是人和动物的一种本能。人在工作、学习、生活中,都接二
First, from the battery life prediction Speaking of Astronautical use of the battery life expectancy is particularly high, it is self-evident matter. However, sealed Ni-Cd batteries, which are a type of coin cell battery, have a long life expectancy when used under stress and can double the difference between similar products. Therefore, the need to speed up the destruction by sampling test to test the quality. Accelerated test is divided into eight individual factors. How to correlate or predict battery life from these eight sets of data? Burton, such as the University of the United States, uses algorithms such as Pattern Recognition to successfully solve this issue. Pattern recognition, also known as “graphics recognition.” The original “graphic recognition” is a human and animal instincts. People in the work, study, life, all take two