瞬态机电耦合作用下的伺服机构动态特性分析与参数优化

The transient response and disturbance rejection capability of servo mechanisms is the important foundation of high precision and efficiency automatic processing. The attention of this project is focused on the phenomenon of transient electromechanical coupling, and the techniques of dynamic characteristics analysis and parameter optimization will be analyzed. The transient parameter coupling model, which involves the input current of servo drive end and the vibration response of mechanical actuator, will be constructed to describe the law of energy transformation and information transfer during the transient motion. The demodulation method for the transient current signal and the parameterization method for the transient vibration signal will be studied to achieve a cross decoupling strategy for electromechanical information. The servo mechanism will be simplified as a nonlinear system, whose excitation is the current inputted into the permanent magnet synchronous motor and response is the vibration appeared to the mechanical structure. The generalized correlation between the excitation and response will be disclosed by the kernel canonical correlation method. The response will be decomposed as the ideal response component and the disturbance response components by an improved empirical mode decomposition method. Then, the generalized delay time, natural frequency, damp, modes and nonlinear disturbance frequency response can be estimated. For enhancing the stability of servo mechanism in transient motion, an optimization method based on the artificial immunity algorithm will be proposed to provide an optimal combination of electromechanical parameters. The main principle of the project is to identify the transient characteristics of generalized nonlinear system. The core content of the project is the decoupling of the time-varying electromechanical information. The research results may provide the foundation to the mechanical and electrical transient matching design. It is significant to push forward the high precision and efficiency development of industrial automation.

瞬态过程中的响应与扰动抑制能力是伺服机构完成高精高效加工的重要保障，本项目以瞬态机电耦合效应为对象，研究伺服机构动态特性分析与机电参数优化技术。建立从伺服驱动端输入电流到机械执行端振动响应的机电耦合模型，描述瞬态机电能量转换与信息传递规律；研究瞬态电流信号的解调分析方法与瞬态振动响应信号的参数化分析方法，建立瞬态机电信息的双向交叉解耦机制；揭示驱动电机电流输入激励与机械结构振动响应之间的广义相关性，构造理想瞬态激励响应与扰动激励响应的自适应分离新方法，实现伺服机构瞬态特性的定性与定量分析，探索以瞬态稳定性为目标的机电参数智能优化方法。其基本原理是从伺服机构瞬态动力输入与响应输出的表征中认识非线性系统特性，其核心内容是瞬态条件下的时变机电信息解耦。本项目的研究将填补伺服机构瞬态特性分析的空白，是工业自动化装置动态性能匹配设计的重要支撑技术，对推动伺服机构向高精度和高效率发展具有重要的意义。

瞬态过程中的响应与扰动抑制能力是伺服机构完成高精高效加工的重要保障，本项目以瞬态机电耦合效应为对象，研究了伺服机构动态特性分析与机电参数优化技术。主要完成工作包括：构建伺服机构模拟实验平台与机电动态信号测试分析系统；建立了包含交流永磁同步电机模型和典型机械执行单元模型的伺服机构瞬态机电耦合仿真模型，完成伺服系统瞬态运行仿真分析；研究多分量调频-调幅信号的精确解调技术，设计了瞬态电流信号的解调分析方法与瞬态振动响应信号的参数化分析方法；提出了基于电流瞬时幅值样本熵与电流瞬时频率峰-峰值的机械执行单元性能评价指标，使用集成经验模态分解方法提取振动响应中的电流谐波扰动分量，利用核规范变量分析法量化机电耦合作用，综合评价伺服机构瞬态运行性能，形成了利用电流特征描述机械特性和利用振动特征描述驱动特性的交叉信息解耦机制；对机电耦合模拟实验平台进行在环仿真改造，提高了数控单元应对复杂优化问题的能力，开展控制与多物理域有限元联合仿真的电机瞬态振动分析，为降低伺服机构瞬态振动提供了理论基础，提出面向伺服系统的模糊自适应深度强化学习方法，利用异步优势演员-评论家算法优化控制参数，提高了伺服系统瞬态运行的响应快速性和稳定性。本项目先后在国内外期刊及会议中发表论文11篇，其中SCI检索4篇，EI检索7篇；申请发明专利7项，其中已授权5项；项目组成员参加境外学术会议3人次，开展境外学术合作1人次。项目开展期间培养硕士研究生6名，博士研究生1人，其中已获得学位5人。本项目的研究是工业自动化装置动态性能匹配设计的重要支撑技术，有利于推动伺服机构向高精度和高效率发展。