化工过程历史数据压缩方法研究

Research on method for compressing recorded chemical process data

针对化工过程的历史数据 ,给出了一种基于信号多分辨分解和小波变换技术的可对任意长度数据进行压缩的算法 ,详细描述了实现原理和计算步骤 .并对一造纸厂的温度、液位、流量及变化较大的压力等物理参数进行数据压缩和还原 ,仿真结果表明 ,在压缩比为 10左右时 ,压缩后的重现数据仍保持较小的相对误差 ,得到了满意的压缩效果 (<0 .5 % ) ,极大地减少了数据的存储空间 .

A compression algorithm based on multi-resolution signal decomposition and fast wavelet transform for arbitrarily long data on chemical process was developed in this work. The method's implementation and computing steps are described. Satisfactory outcome was obtained when the method was applied to the compression decompression process of temperature, liquid level, flux and various pressure signals from a paper mill. The simulation results showed that the compression ratio should optimally be about 10, and when the storage requirement is reduced the error is still and small enough to accept.