基于学习模型的刀具磨损少传感识别及表面状态预测与控制方法

Ni-based super alloy blisk and ring type casing are the critical components in aero-engine, and their machined surface state will determine the service performance and safety of the aero-engine. This project takes the tool wear identification and its evolution in the time-varying milling process of GH4169 blisk and casing as the breakthrough point, and the establishment of the machined surface state control while considering the tool wear as the target. Based on multi-sensing environment, the relationship among the milling condition, tool wear and the corresponding machined surface state will be established, and learning models for tool wear and machined surface macro-micro state prediction will be developed. Furthermore, a novel tool wear identification method under sensing environment with few sensors will be studied, the evolution of tool wear and machined surface state under varying milling conditions will be revealed. Based on that, the on-line correction as well as the machined surface quality determination methods will be developed. Taking the tool wear and machined surface state as the criteria, this project will develop the optimization methods for the process parameters, machining area division and tool usage control, thus to control the machined surface state and to ensure the surface quality consistency for the critical components. Based the research results of the project, breakthroughs in both the theory and key technologies for the machined surface state control of high value-added critical components will be made, thus to enable the high performance manufacturing of the aero-engine.

高温合金材料整体叶盘、环形机匣是航空发动机的关键构件，其加工表面状态对发动机的服役性能和安全起决定性作用。本项目以GH4169材料整体叶盘和环形机匣关键构件时变铣削过程中的刀具磨损识别及其演化为突破口，以建立考虑刀具磨损的铣削加工表面状态控制方法为目标，基于多传感试验环境研究铣削工况-刀具磨损-表面状态的关联映射关系，建立刀具磨损及加工表面宏微状态预测的学习模型；探索现场刀具磨损少传感识别方法，揭示时变铣削过程的刀具磨损及表面状态演化机制，建立预测模型的在线修正与加工过程质量判定方法。以刀具磨损及表面状态为准则，研究考虑刀具磨损的高温合金关键构件加工工艺参数域、加工区域划分及刀具使用优化方法，实现关键构件加工表面状态的控制，保障加工表面质量的一致性。在航空发动机高温合金材料关键构件铣削表面状态控制理论和关键技术上取得突破，为实现航空发动机的高性能制造提供科学支撑。

高温合金材料整体叶盘、环形机匣是航空发动机的关键构件，其加工表面状态对发动机的服役性能和安全起决定性作用。针对高温合金切削加工刀具磨损现象严重，加工表面质量难控制等问题，本项目以GH4169材料切削加工过程为研究对象，重点开展了刀具磨损与工件表面质量演化分析、刀具磨损预测方法、加工表面质量预测方法与考虑刀具磨损的加工表面质量控制等研究。在刀具磨损演化方面，分析了监测信号与刀具磨损的关联特征，建立了刀具磨损监测信号的聚类演化分析方法，间接获得了刀具磨损的多阶段动态分布特性。在刀具磨损离线预测方面，建立了考虑刀具切削刃位置的磨损速率模型，实现了刀具磨损沿切削刃分布的预测，并开发了刀具磨损预测模块。在刀具磨损在机预测方面，面向加工现场数据采集难的问题，建立了少传感刀具磨损预测方法，实现了仅利用切削力信号在机识别刀具磨损情况。在加工表面质量预测方面，综合考虑多个刀具磨损形式对表面质量的影响，建立了加工表面粗糙度预测的神经网络。在加工表面质量控制方面，建立了加工工艺参数优化数学模型，以预期加工质量为约束，随着切削加工进行根据刀具磨损情况动态调整加工参数，实现了加工表面质量的一致性。研究内容在镍基高温合金材料自由曲面及凸台特征的加工中进行了验证，成果可在航空发动机整体叶盘、机匣等零件的加工中进行推广应用。