基于相似性匹配的机泵监测模拟量数据压缩方法

Analog Data Compression Method Based on Similarity Matching for Pump Monitoring

为了解决关键机泵在线监测过程中模拟量数据存在的难压缩问题,提出一种基于相似性匹配的改进旋转门算法。通过构建监测过程模拟量数据的实时状态模型和历史状态矩阵,提出一种基于窗口平移的状态模型相似度量方法;针对传统旋转门算法门限值难以确定、门限值单一的问题,提出动态门限值计算模型;在上述研究的基础上,通过相似性匹配动态调节压缩过程中门限值的大小,研究了改进旋转门算法的原理、步骤及解压缩方法。最后提出了算法的性能评价指标并且采用真实的监测数据对算法的压缩性能进行了测试,验证了改进旋转门算法比传统旋转门算法具有更好的压缩性能。

To solve the problem of difficult compression of analog data in on-line monitoring of key pumps,an improved SDT algorithm based on similarity matching is proposed.By constructing the real-time state models and the historical state matrix of analog data,a similarity measurement method of state models based on window translation is proposed.The threshold value of traditional SDT algorithm is single and difficult to determine.On the basis of the above research,a dynamic threshold calculation model is proposed to dynamically adjust the threshold value by similarity matching.The principle,steps and decompression method of improved SDT algorithm are studied.Finally,the performance evaluation index of the algorithm is proposed and the compression performance of the algorithm is tested with real monitoring data,which verifies that the improved SDT algorithm has better compression performance than the traditional SDT algorithm.