Digital arc welding power supply based on real-time operating system

A digital arc welding power supply was designed with the advanced reduced instruction set computer machine （ARM） and embedded real-time multi-task operating system micro C/OS- Ⅱ. The ARM, with its powerful calculating speed and complete peripheral equipments, is very suitable to work as the controller of the digital power supply. The micro C/OS- Ⅱ transplanted in ARM, helps to improve the respondent speed against various welding signals, as well as the reliability of the controlling software. The welding process consists of nine tasks. The tasks of great significance on reliability of the welder, for example, the A/D conversion of current and voltage, enjoy top priority. To avoid simultaneous-sharing on A/D converter and LCD module, two semaphores are introduced in to ensure the smooth performance of the welding power supply. Proven by experiments ,the ARM and the micro C/OS- Ⅱ can greatly improve both the respondent speed and the reliability of the digital welder.     

机器人CO2焊接动态过程的分析

在详细分析影响机器人CO2焊等速送丝系统稳定性各个变量之间的动态关系基础上,建立了系统的动态模型.以此模型为基础,进行了稳定性仿真、抗扰性仿真、动态调节仿真和系统改选分析.     

Comparison of the pulsed MIG welding process for different median current

Based on the high speed video system with electrical signals collecting and wavelet analyzing, the welding processes under three different median currents with the same median time were detected. The experimental results show that, when the median current is higher, the input peak energy is lower. And the droplet transfer is almost in spray mode. The welding process is the stablest.     

软开关逆变点焊控制电路的设计

通过对逆变电阻点焊设备的现状分析，介绍了软开关逆变点焊的控制系统．讨论了主控系统、反馈系统的工作原理．并具体设计、实现了单片机对移相芯片UC3879的高精度控制，电流反馈的控制等。     

焊接过程的神经网络建模及控制研究

采用人工神经网络方法对焊接过程进行建模及控制，并以ＧＴＡＷ作为对象加以实施。用误差反传网络（ＢＰ网）建立了静态过程ＧＴＡＷ熔透情况下表征其质量的正面熔宽、背面熔宽与施焊参数之间的关系，即静态模型，并从质和量两个方面验证了该模型的可行性。提出了采用两个网络对ＧＴＡＷ动态过程建模及控制的方法，并通过实验验证了所提方案的正确性。文中所提的方法从根本上不同于以往的以固定结构数学模型为基础的建模及控制的方法，无需对过程作任何假设，所建模型与实际过程符合较好。实验证明，该方法具有容错性好、通用性强和抗干扰能力强等优点，为焊接过程的建模及智能化控制找出了一条新途径。     

A study on frequency domain mathematics model of arc welding reverse converter

Arc welding converter (AWC) power source is a nonlinear and discrete a bnormal system. It is difficult to get the accurate analytic solution in theory. In this paper, an average state equation of the system is established by means of state space analysis method. The mathematics model of the reverse converter is gotten after the average equation of the system is linearized. The model is a two ranks oscillation link. With the parameters of the AWC, when the condition of (1/RC)2 >>4/LC is available, the mathematics model may be simplified to as erial structure which includes a integral link and a one rank inertia link. On the basis of this, it can be analyzed and rectified in frequency domain, then some groups of reality rectification parameters are gotten around the theory rectific ation parameters. These groups of parameters are taken to do the resistance load experiment and welding technology experiment. The result of experiment accords with the theory analysis result. This paper also provides some practical analysis of the model under the large signal process such as the sudden opening of the load, short circuit, turning on and turning off the machine. The result of the research shows that the power source system of AWC can be analyzed and rectified by the frequency domain analysis method in the engineering. The deduced mathemat ics model of small signal average frequency domain shows the basic characteristics of the system, and the structure parameters of the system is formed by R, L, C and switching ratio D. The model is very simple and clear, and it can be used to analyze and design the power source system of AWC and it can also be used to direct the regulation for dynamic characteristics of the system.     

逆变式电阻焊机的研究进展

介绍了逆变电阻点焊机的特点,国内外逆变点焊机的研究现状,提出了逆变点焊电源研制过程中存在的几个问题。     

基于变分的SAR图像目标特征增强方法

研究了SAR图像目标特征增强的变分方法.通过分析P-M扩散方程中的扩散系数,得出梯度(Canny)边缘检测算子对SAR图像的非恒虚警性使得滤波后图像的边缘变得模糊.而ROA算子能有效地检测出图像中的边缘,但较难检测出图像中的强散射点目标.基于SAR幅度图像中相干斑噪声的Rayle igh分布,从最大后验概率估计出发,结合ROA边缘检测图像以及SAR幅度信息来构造扩散系数,建立SAR图像目标特征增强的变分模型.实测SAR图像处理结果显示该方法在充分抑制均匀区域的相干斑噪声的同时能较好地保护并增强图像的边缘和强散射区域.     

基于Lyapunov指数刻画短路过渡GMAW的稳定性

基于不同焊接速度下短路过渡熔化极二氧化碳气体保护电弧焊的焊接电流时间序列,利用Takens相空间重构技术和Lyapunov指数的Wolf算法,分析评估了焊接系统的最大Lyapunov指数。结果表明,当焊接过程的稳定程度较高时,最大Lyapunov指数较小;而当系统处于不稳定状态、状态变动迅速时,最大Lyapunov指数相应较大;因此最大Lyapunov指数和焊接过程稳定性成负相关关系,可以将其作为刻画短路过渡熔化极二氧化碳气体保护电弧焊稳定性的一种新的量化指标。     

基于神经网络的焊接过程建模及控制研究

采用人工神经网络方法对焊接过程进行建模及控制．并以ＧＴＡＷ作为对象加以实施。用误差反传网络（ＢＰ网）建立了静态过程ＧＴＡＷ熔透情况下表征其质量的正面熔宽、背面熔宽与施焊参数之间的关系，即静态模型，并从质和量两个方面验证了该模型的可行性。提出了采用两个网络对ＧＴＡＷ动态过程建模及控制的方法，并通过实验验证了所提方案的正确性。文中所提的方法从根本上不同于以往的以固定结构数学模型为基础的建模及控制的方法，无需对过程作任何假设，所建模型与实际过程符合较好。实验证明，该方法具有容错性好、通用性强和抗干扰能力强等优点，为焊接过程的建模及智能化控制找出了一条新途径。