【作者】 张涛；

【导师】 桂卫华；

【作者基本信息】 中南大学 ， 控制理论与控制工程， 2010， 博士

【摘要】 熔化极气体保护焊(Gas Metal Arc Welding)是一种高效节能的焊接工艺,通过连续送进的焊丝与焊接工件之间产生的电弧作为热源熔化焊丝和金属母材,在焊接过程中外加保护气体避免熔滴、熔池金属及焊接区高温金属受到周围空气的氧化作用,特别适合于自动焊接生产线和机器人焊接系统。GMAW焊接过程熔滴飞溅大,焊缝成形差,焊接参数调整困难,限制了GMAW焊接工艺的推广应用。熔滴过渡过程直接影响GMAW焊接质量,由于焊接规范的不同,熔滴存在多种过渡模式,目前国内外的熔滴过渡控制方法主要从作用于电弧的电流和电压入手,针对特定的熔滴过渡模式,通过对焊丝输入热量的控制减小飞溅,或通过对弧压的控制保证焊接过程的稳定,难以适应不同工艺参数的要求。根据GMAW焊中存在的问题,分析了熔滴过渡过程的机理,针对短路过渡模式和自由过渡模式中熔滴不同的行为状态,采用不同的控制策略保证了熔滴过渡的平稳和连续。在此基础上,设计完成了基于软开关的能适应所有焊接参数和工艺条件的GMAW焊接系统。主要研究工作和成果如下：(1)在对熔滴受力和熔滴的运动过程进行分析的基础上,分析了焊接电流、焊接电源输出电压、焊接速度等工艺规范参数对熔滴过渡模式和焊缝成形的影响,从熔滴位移和质量变化、焊接回路动态过程、弧长变化方面对焊接过程进行了数学描述,建立了GMAW焊接过程数学模型,并对模型进行了仿真,证实了模型的准确性。为基于模型的控制策略提供了理论基础。(2)针对熔滴飞溅大和焊缝成形不好的问题,提出在熔滴短路过渡模式中采用分阶段电流波形控制策略,将熔滴短路过渡的物理过程分成五个阶段,根据熔滴在每个阶段的状态和受力情况,输出不同波形的焊接电流,保证熔滴在短路阶段能以较小的飞溅平稳过渡。建立分阶段电流波形控制的仿真模型,为分阶段电流波形控制参数的优化提供依据。采用模糊神经网络控制算法调节大电流燃弧段的作用时间维持弧长的稳定和熔滴过渡的一致性和平稳性,利用遗传算法对模糊变量的隶属度函数进行调整和优化,将优化后的网络引入GMAW短路过渡模式的焊接过程,通过附加动量法在线调整神经网络的权值,在减小飞溅的基础上,保证焊缝的成形。(3)在熔滴自由过渡模式中,采用直接模型参考自适应控制算法实现焊接参数的去耦控制,解决了GMAW焊接参数调节困难的问题。直接模型参考自适应基于指令跟踪器和输出渐进跟踪理论,控制器的设计只需利用参考模型和跟踪误差来完成,与控制对象几乎无关,控制结构简单,可实现高阶被控对象对低阶参考模型的性能跟踪。针对目前离散时间系统直接模型参考自适应控制算法中存在的稳态误差和控制律计算困难的问题,提出改进算法消除稳态误差,且控制律的计算符合因果关系。分析了算法的跟踪效果和稳定性,采用该改进算法,通过调节焊接电源的输出电压和送丝速度,实现焊接电流的和弧长的稳定,保证熔滴尺寸和过渡时间间隔均匀一致,焊接过程平稳流畅。(4)为了提高焊接电源动态性能,提出一种适合IGBT焊接电源的零电压零电流软开关拓扑结构,通过在中频变压器原侧增加一个谐振电容和饱和电感,实现了超前臂的零电压开通和滞后臂的零电流开通和关断。对超前臂和滞后臂上的开关管的开通和关断损耗进行了分析,并提出了减少超前臂开通损耗的途径。针对焊接电源在GMAW焊接过程中负载变化剧烈,提出在高频变压器副侧增加一个换流电感,保证空载和轻载时软开关条件的充分实现。根据软开关拓扑结构,设计了焊接电源,实验结果验证了理论分析和设计。(5)采用模块化设计思想,开发了适合全范围焊接参数,实现多过渡模式下熔滴过渡控制的GMAW焊接系统,系统由80C196KC单片机控制的基于软开关技术的GMAW焊接电源模块和GMAW送丝系统模块组成,两个单元之间采用通信协议进行数据传输和交换。焊接系统具有完备的过热、过流、欠压保护措施和软硬件抗干扰措施。现场焊接试验结果表明,系统较好地解决了焊接过程飞溅大和焊接参数的匹配范围窄的问题,保证焊缝成形和提高生产效率,使GMAW焊工艺性能得到全面的提高,具有重要的实用价值。更多还原

【Abstract】 Gas metal arc welding(GMAW) is suitable for the automatic welding and robot with high efficiency and energy saving. The continuously supplied wire and welding pool are heated and melt by the arc between them to form a welding seam. In the welding process, the welding area are protect by the shielding gas from the hazard action of the air.Spatter, poor appearance of welding seam and hard-matched welding parameters are main factors which impact on the promotion and application of GMAW. The welding quality is related to the droplet transfer methods. The GMAW process provides short circuit transfer and free transfer methods. The currently methods of control of droplet transfer is to eliminate the energy to droplet to reduce the spatter and keep the arc voltage to make the welding process steady. It is difficult to be suitable for different droplet transfer methods and welding parameters.The scheme of a power source with different control Strategies adapt to different droplet methods are designed to make the process of transfer steady and get a high quality of welding seam with lower spatter. The main research contributions of the thesis are as follows:(1)Based on the analysis of the forces acting on the droplet and the transfer process, a mathematical model describing the GMAW process is developed. The mathematical model includes a description of the droplet displacement and mass, electrical circuit, the arc length, and the melting rate. Simulation programs show the effectiveness of the model, based on which an adaptive contol algorithm is developed.(2)In the short circuit transfer mode, the droplet transfer process is divided into five periods, the corresponding current wave form that adapt to the states of the droplet in each period is output. A big current is applied on the period of arc burning to provide the enough energy to ensure the good appearance of welding seam. The principle of five periods wave form is dicussed and the simulation model of wave form droplet transfer control is developed to optimize the parameters of wave form. The duration of arc burning period is adjusted by the fuzzy neural network algorithm to control the arc length to the preset value. The membership function of fuzzy variable is optimized by experiment data with genetic algorithm. On the process of welding, the additional momentum method is used to adjust the weight of the neural network on the realtime.(3)To solving the problem of hard-matching of parameters in free transfer mode, the simple direct model reference adaptive control is applied to decouple the welding parameters. The simple direct model reference adaptive control algorithm is based on the the theory of command generator trace, Only the states of the reference model and the trace error are needed to design the controller that let the high order plant model trace the simple chosen low order reference model. A improved discrete simple direct model reference adaptive control algorithm is proposed to remove the steady state error. Current and arc length as being the process output are controlled by open circuit voltage and wire feed speed to trace the output of the reference model, which ensure the size and the transfer interval is uniform.(4)The application of droplet transfer control strategies depend on the rapid dynamic characteristics of the welding power source. The performance of the traditional welding power source is hard to satified the requirements. The soft switch technique is applied to improve the switching frequency of inverter which is the key to the dynamic response of the power source. A zero voltage and zero current full bridge soft switching topology is proposed, in which a dc blocking capacitor and a small saturable inductor is added in the primary side to achieve ZCS for the lagging-leg switches.The working states of soft switching circuit is analyzed. The way to eliminate the turn-on switching loss of the leading-leg switches is discussed. The conditions that achieve the soft switch in a wide range of load which is from off-duty to full duty in the welding process is hard to be satisfied, so a inductor is add on the secondary side to provide the additional energy. A power source based on the topology is designded and the successful tests have been carried out.(5)A welding system suiable for the different parameters and droplet transfer method is developed based on the module design. The system are consist of microprocessor controlled welding power source unit and wire feeder unit which exchang the welding parameters by communication. The complete protection of overheat, over current and undervoltage are included.The welding experiments show that the system can get high welding quality with lower spatter. The droplet transfer control methods and the system are valuable for the application of GMAW.更多还原