Research On Pipeline Magnetic Flux Leakage Inspection Data Compression Technology

In order to enhance the measure precision and expand the using area of the existing oil pipeline MFL(Magnetic Flux Leakage) detecting system to natural gas pipeline testing, detailed information of flux obtained by decreasing sampling time and space interval is needed. The huge amount of data produced by this will exceed the processing ability. Data compression is an effective way of solving this problem. The data compression algorithm must be fit to hardware implementation for the new pipeline magnetic flux detecting system will adopt FPGA (Field Programmable Gates Array) as the system kernel. This paper, combined with the task and based on the thorough analyses of the theories and methods of data compression, realize the data compression of the MFL inspection data using Huffman coding and wavelet transform coding. Those compressions algorithms are suited to implementation by hardware and have good realtime quality because they don't have multiplication and division operation. Those algorithms can achieve compression ration over 10 to 1 at the same time of keeping all the important detecting correlative data by using lossless method to compress important data and using lossy method to compress insignificant data.At the exordium, this paper introduces the task's origin and significance and introduces the research situation on pipeline MFL data compression at home and at abroad. The second chapter explains the character of MFL data by introducing the principle of pipeline MFL inspection and the structure of MFL detector, and then brings forward the diagnostically lossless compression request by combine discuss the analyzing method of inspection results. The third chapter briefly introduces the theory and methods of data compression. The fourth chapter discusses the Huffman coding method, anticipating coding method and wavelet transform coding method used in this paper. The fifth chapter use Huffman coding method of the row difference data and lifting wavelet transform and SPIHT (Set Partitioning in Hierarchical Tree) coding method to realize the diagnostically compression of MFL data by judging the significance of data using changing threshold and dynamic area threshold of MFL data. Thesixth chapter introduces the structure of FPGA and design flow at first and then introduces the related content of the hardware implementation of compression algorithms. Finally, The conclusion part summarizes the whole work of the paper.