铣削加工过程多源能效状态在线识别方法研究

It is essential in green manufacturing that machining systems are always operating in energy saving states, which can only be evaluated by real time identification of energy efficiency state of the system online. The objective of research efforts in this project is to meet this requirement. The work starts with the exploration of the characteristics, signal features and their coupling and mixing mechanisms of energy efficiency states in monitoring signals of a machine system. And proper signal feature decoupling and separating algorithm will also be developed. Consequently, the identification and source tracking of inefficient energy consumption conditions in a machine system will be studied. As a matter of fact, all parts of a machining system affect the general energy efficiency state of the system, so identification of energy efficiency states is a multi source recognition problem. The suggested solution to this problem in this project is, based on the attributes of energy consumption of the sources, to define those states into direct energy efficiency state, indirect energy efficiency state, and hybrid energy efficiency state whose signal features will be extracted, separated, and tracked from the monitoring signals. The relationship as well as differences among energy efficiency state, tool condition and machine condition will also be investigated. Finally, the tracking algorithm will be explored to track energy inefficiency conditions and identify energy efficiency faults to provide theoretical foundation and potential technical solutions for online monitoring of energy efficiency states of machining systems.

绿色制造要求切削加工系统总是运行在节能状态，而对运行中的切削系统，判断其是否节能，需要对其能效状态进行实时在线识别。本项目的研究就是围绕这一需求展开的。项目从研究监测信号中反映切削系统能效状态的信号属性、特征及其耦合和叠加机理入手，探索相应的信号特征解耦或分离算法，进而解决切削系统低能效状态的识别与溯源问题。事实上，切削过程的各个环节都可能造成系统总体能效低下，因此，切削过程能效状态识别是一个多源状态识别问题。为此，本项目根据切削能耗属性，将切削加工系统的多源能效状态，定义为直接能效、间接能效及复合能效状态。然后，分别研究监测信号中这三种能效状态信息的提取、分离及溯源方法，同时探索并揭示能效状态与刀具状态、切削状态及设备状态等传统监测目标状态的内在联系及区别。最终，通过溯源算法研究解决能效状态的溯源及低能效故障状态的定位问题，为切削加工系统的能效状态在线监测提供理论依据和潜在技术解决方案。

绿色制造系统期望运行在节能状态，能效状态的在线识别是判断制造系统是否处于节能状态的关键。本文探索采用在线监测的方法识别铣削过程能效状态。首先，明确铣削过程能效状态在线监测所需获取的物理量。然后，基于试验对相关物理量进行多源传感器数字信号和温度场热度图像信号采集。设计多源耦合传感器信号的解耦与分离算法，提取铣削过程能效状态特征；同时，研究铣削过程温度场热度图像的能效状态聚类分析方法和特征提取方法，获得温度场热度图像能效状态特征。再次，提出铣削过程多源耦合能效状态识别算法，建立能效状态特征集，对铣削过程目标状态和耦合状态进行综合识别。最后，研究铣削过程低能效状态溯源与定位问题，建立低能效状态溯源与定位策略，探索特征选择排序方法，通过设计切削系统单一源头和复合源头导致低能效状态的铣削试验，确定切削系统组成要素导致低能效状态的来源。铣削过程能效状态识别方法的研究有望解决制造系统能效状态在线监测问题和低能效状态溯源问题，为建立绿色制造节能策略提供评估依据与手段。