摘要
根据多孔算法思想和应用Matlab小波工具箱对实际焊接电压信号的分析,选择db2小波基,提出了一种适合实际CO2焊接过程短路信号的小波检测算法.该算法使滤波器序列静止不动,采样信号序列动态的前进,对输入信号采样一次,处理一次,由于检测过程对时间要求非常苛刻,且小波算法包含大量的乘和累加运算,运算量很大,故选择DSP(TMS320LF2407A)为处理器,采用C++编制主程序;同时充分利用DSP的硬件乘法器和RPT指令,用汇编语言编制了计算量小、执行速度快的小波变换子程序;将该子程序嵌入主程序中,采用C++与汇编混合编译的方式对采样信号进行三级分解来检测短路信号的发生时刻.试验表明,算法的正确性,并能实时、准确的检测出真短路信号的发生时刻,满足波形控制要求.
According to algorithm a'trous and analysis of CO2 arc voltage characteristic by Wavelet Toolbox in Matlab,the wavelet algorithm which is suitable for detection of short-circuit signal in practical CO2 welding process is derived.Different wavelet basic waves are used to analyze the CO2 arc voltage signal in Matlab,finally the db2 basic wave is selected and used in this algorithm in which the filter is stationary and the sampling signal is processed one by one by three-decomposition.To meet real-time requirements,the main program is developed in C++ and the assembly subroutine of wavelet detection is developed by the multiplier of DSP and the RPT instruction.As a result,it has a fast operation speed and improves the real time capability of detection.The calculation results of C++ program show the accuracy and the feasibility of this algorithm.Experiment results show that false short-circuit signal can be removed by the db2 basic wave and the short-circuit moment can be decteted in real time by this algorithm.
出处
《焊接学报》
EI
CAS
CSCD
北大核心
2010年第1期93-96,共4页
Transactions of The China Welding Institution
关键词
小波变换
短路信号检测
算法
实时性
wavelet transform
detection of short-circuit signal
algorithm
real-time
