一种基于提升格式的过程数据压缩新方法

为了有效利用数据压缩中过程数据的局部特征，提出了一种基于提升格式的自适应压缩方法.该方法根据信号的局部特性选择预测环节，利用塔式分解中小波系数不参与下一层分解的特性，对小波系数做简单的可逆变换，将选择预测环节的信息保存在小波系数中.理论分析及仿真结果均表明，预测环节的选择信息不需要额外的存储空间，不会由于阈值处理和压缩编码受到损害，保证了逆变换的稳定性.与非自适应的插值小波以及常用的Haar小波相比，这种自适应的提升格式降低了均方根误差(RMSE)和单点误差(LPE).

Novel method of process data compression based on lifting scheme

To efficiently use the local character of process data in data compression,an adaptive compression method based on lifting scheme was proposed.The prediction step of lifting scheme according to the compression data was adaptively adjusted in the method.The wavelet coefficients served as a simple reversible transform,and stored the sign of choosing the prediction step by using the properties of pyramid algorithm.Theoretical analysis and simulation results confirm that the compression encoder and wavelet shrinkage do not affect the sign of choosing the prediction step and that the inverse transform is stable.The adaptive lifting scheme performs better than the interpolating wavelet and Haar wavelet,and can reduce the root mean square error(RMSE) and local point error(LPE) without the redundant storing space of the sign.