大规模向量式有限元行为数据无损压缩模型

Massive Vector Form Intrinsic Finite Element Behavior Data Lossless Compression Model

摘要:

针对向量式有限元产生的海量的高精度行为数据在基于web的体系架构的数据传输和文件读取的瓶颈，依据向量式有限元行为数据的内部联系和冗余特点，建立了行为数据的数学模型，并对冗余数据进行删减和合并，提出一种有效的无损压缩模型.最后，给出基于模型的压缩算法.经过验证，数据的规模显著减小，极大提升了数据处理的效率.解压时则完全还原原始数据，完全不影响行为数据的精度.

Abstract:

The behavior data of vector form intrinsic finite element computing has a large size as well as high accuracy, which become the bottleneck of data transmission and read that based on web. In order to resolve the problem. According to the characteristics of redundancy and the internal relations of behavior data, a mathematic model for the behavior data is proposed and according to the model, a lossless compression model is put forward by deleting and merging the redundancy data. In the end, an efficient compression algorithm is given based on the model. Experiment indicates that the algorithm greatly compresses size of behavior data and accelerates the load speed of data. The decompressed file can be restored to the original data and the accuracy of behavior data doesn’t be affected at all.