大规模向量式有限元行为数据压缩模型及算法

Massive Vector Form Intrinsic Finite Element Behavior Data Compression Model and Algorithm

摘要:

针对向量式有限元程式计算所产生的结果数据体积庞大，上传云端和读取处理的速度瓶颈，必须对其进行有效的压缩.通过对向量式有限元行为数据进行分析，依据数据本身的冗余度高、无效信息多、数据格式较长等特点，进而提出一种适用的压缩算法，对行为数据进行时间帧合并、删减冗余数据、舍去过长精度、线元素增量表示等压缩方法.经过验证，本算法压缩后的数据体积明显减小，而程序读取数据的速度则有显著提高，大幅改善了向量式有限元行为分析软件的使用效率.

Abstract:

The result data of the mid processing computation of vector form intrinsic finite element analysis has a large size. That is a considerable bottleneck in the speed of the Cloud uploads and program load, which should be highly compressed. The behavior data of the vector form intrinsic finite element is studied. After analyzing the characteristics of high redundancy, invalid data and numerical notation format of the data, a practical compression algorithm is proposed to merge the time information, delete redundant data, round the over precision and use the line element incremental expression. The numerical experiments indicate that, this algorithm highly compresses the behavior data size and accelerate the data load speed. It greatly improves the usage efficiency of the vector form intrinsic finite element behavior analysis software.