Study And Design Of Yarn Signal Spectrum Analysis Filters Based On FPGA

Yarn unevenness is an important indicator of measuring yarn quality, also is the decisivefactor affecting the appearance of cloth, gauze so the detection of yarn unevenness is theimportant tool in the process of spinning controling the quality effectivily. Textile industry isusually using spectrum analysis method for the detection of yarn unevenness, the essence ofwavelength spectrum analysis is the frequency spectrum analysis. In the textile industry site,the spectrum of yarn signal is often realized using the filtering method. Traditional spectrumanalysis of yarn signal is achieved by analog filtering method, the essence of which method isachieved by setting up multiple sets of analog filters.But this method has many shortcomings,such as poor anti-interference as well as complex circuit, and thus can’t guarantee thereliability and stability of the spectrum analysis.With the rapid development of the digital signal processing technology, the digital filtersare widely applied in the spectrum analysis, voice and image processing and other fields. Thecurrent implementation methods of digital filter mainly include two categories:one is usingDSP chip，the other is using the PLD device (including FPGA and CPLD).There are someproblems of implementation method based on DSP processor exist,such as the algorithmscomplicated, development and maintenance difficulties and the real-time poor. Although themethod has overcomed the complexity of analog filtering circuit, poor anti-interference, andso on.But if using the PLD device to achieve filters, you can use the hardware multiply-addmodules,thus the computing speed is much faster than using DSP devices. Therefore, theimplementation of the filters based on the self-programming way of PLD has become thepreferred chooce.In summary,the article propose to use FPGA device to design multi-channel digital filtersfor spectrum analysis of yarn signal, on the one hand because of the FPGA device can run inparallelly, thereby reducing the run-cycle of the program required to achieve filteringalgorithm, ensure real-time of the spectrum analysis; on the other hand, the FPGA device hasthe charactors: application portability, convenient and flexible interface configuration, etc.Sothe spectrum analysis filtering algorithm which is implemented based on FPGA can beflexibly transplanted and configurated in different FPGA devices. First, this paper describes the importance of spectrum analysis of yarn signal, as well asthe basic principle of spectrum analysis filtering algorithm and algorithm optimization theory,and proposes using FPGA device to achieve the algorithm, in order to realize multi-channeldigital filters for spectrum analysising of yarn signal. Second, using MATLAB software tocomplete the design of the spectrum analysis algorithm and optimization algorithm, simulatesand verificates the effect and performance of filters designed,the results confirms therationality and feasibility of the spectrum analysis filtering algorithm as well as optimizationalgorithm;the implementation process of the spectrum analysis filtering algorithm as well asoptimization algorithm based on FPGA has been simulated using MATLAB software,which can confirm using FPGA devices to implement algorithm is feasible; andmulti-channel spectrum analysis filters have been achieved in the MATLAB software. Finally,using the Quartus Ⅱ software designs and simulates the spectrum analysis algorithm of yarnsignal as well as optimization algorithm, and the simulation result based on FPGA arecompared to the theoretical simulation results on MATLAB software, the results are theconsistent, which verifies the feasibility and correctness of the algorithm based on FPGA，andin line with the design mind.