【作者】 蒋永翔；

【导师】 任成祖； 王太勇；

【作者基本信息】 天津大学 ， 机械制造及自动化， 2010， 博士

【摘要】 随着先进制造技术领域对制造质量要求的提高,加工过程故障辨识及控制技术凸显其重要性,对铣削无颤振加工稳定性的在线监测是其重要研究内容之一。本课题以TDNC-H8数控机床为验证平台,进行了稳定性快速预报和在线寻优控制理论的研究,探讨了确保稳定铣削的在线监测智能控制一体化解决方案。基于离散动力系统非线性控制理论,据C-C算法获得的最佳嵌入维和时间延迟进行了在线监测振动信号的相空间重构,通过对可反映颤振规律和可预言性的粗粒化熵率特征提取及实验验证,实现了铣削不稳定加工状态的快速准确预报。考虑多自由度和时变铣削力系数的影响,对厚壁铣削件进行了加工稳定性预测研究。以时变铣削力系数下的铣削力建模为基础,建立了适合于厚壁铣削件的稳定性动力学模型,进行了多自由度耦合及考虑时变铣削力系数的三维稳定性图绘制。通过各加工参数n、ap、ae和fz的稳定性影响分析及实验验证,获得了适合于厚壁件铣削过程的稳定性预测理论。在铣削力系数时变稳定性预测理论研究基础上,探讨铣削至薄壁件后刀具—工件系统刚度时变对稳定性的影响,建立了刀具—工件系统耦合传递函数模型,研究了适合于在线监测的运行模态主轴转速扫频传递函数实时获取方法,构建了准确性、实时性均优的加工稳定性预测理论。研究了以不稳定加工状态快速预报理论进行识别,同时考虑生产规划和加工阶段最大材料去除率的稳定性在线寻优控制方法。探讨了生产规划阶段影响稳定性区域面积的相关结构参数变化规律及加工阶段最大材料去除率获取方法,研究了在线寻优控制中扩大稳定性区域和稳定区快速搜索方法的具体实施。构建了基于网络,具有可重构和开放性的功能模块化嵌入式在线监测智能测控系统。重点针对笔者开发的软件部分进行了嵌入式在线监测单元的开发及功能阐述。建立了加工稳定性在线监测寻优控制及其他多种在线监测算法的应用平台,为本课题所研究算法的工程应用奠定了基础。更多还原

【Abstract】 With the demand of high machining quality in advanced manufacturing field, fault identification and control become the highlighted important technology in machining, and online chatter detecting of milling stability is one of the research contents. Take TDNC-H8 CNC machine as the verification platform, the instability quick forecast theory and stability online optimizing control was researched, the integration settlement scheme of stability online detecting was studied to make sure the chatter free as well as high precision and efficiency.Based on the nonlinear control theory in discrete dynamic system, the C-C algorithm was used to determine appropriate embedding dimension and time delay. The coarse grained entropy rate feature was extracted to reflect the predictable rule in chatter. Through the validation experimenst, the CER can be applied in instability forecast quickly and precisely.Considering the effect of multi-degree and time-varying milling force coefficients, the milling stability prediction theory in thick-wall milling process was studied. Based on the milling force model with time-varying milling force coefficients, the thick-wall milling stability dynamic model was discussed, which lead to the three dimension stability lobe chart with multi-degree coupling and time-varying milling force coefficients. By analyzing the stability state with milling parameters of n, ap, ae , fz as well as milling validate experiment, the stablity prediction theory which is sutiable to thick-wall milling was setup.Based on the time-varying milling force coefficients stability prediction theory, the stability of tool-workpiece system considering stiffness time-varying in thin-walled milling was studied. By modeling the tool-workpiece coupling system transfer function, the online real-time acquisition of transfer function method using spindle frequency sweeping of operational modal was discussed, the chatter prediction theory with high precision and real-time was setup.Taking the instability forecast theory for chatter detecting and by considering maximum materials removal rate (MRR) of production planning and machining stage, the online optimizing control method was studied. In production planning stage, the structure parameters to enlarge stability zone and the MRR acquisition method in machining stage were discussed, and the solution to enlarge stability zone and fast searching online optimizing control was implemented.Based on network, the modularization embedded online detecting system with the characteristic of reconfiguration and open system architecture was studied. Focused on the author’s main work of software programming, the software design methods were discussed. The online detecting experiment validation platform can be used by programming several kinds of algorithms, which takes the foundation of engineering application to the chatter online detecting and control theory.更多还原