摘要
针对数控机床加工过程中的不可控制参数变化的问题,提出一种基于BP神经网络的球头铣刀恒力切削控制方法。分别建立铣削过程模型与BP神经网络控制模型,针对多指标变化导致恒力控制球头铣刀切削难以实施的问题,利用多层BP神经网络的自主高度拟合特性,结合恒力控制多时刻切削力与相关伺服输入,实现自适应调节多指标完成恒力控制。其中,指标包含切削力、进给量、进给速度、背吃刀量。为进一步探究噪声对控制算法鲁棒性的影响,利用方波噪声对所建立的控制算法进行干扰实验,讨论控制算法的抗干扰能力。结果表明:所提出的控制方法能够对球头铣刀控制进行快速且精确的恒力跟踪,在切削深度加深后能自适应地调整切削力;在添加噪声的恒力控制中,所获得的仿真曲线与未添加噪声时保持一致,进一步证明了该方法的抗干扰能力。该控制方法具有较好的控制性能与鲁棒性。
Aiming at the problem of uncontrollable parameter change in NC machining process,a constant force cutting control method of ball end milling cutter based on BP neural network is proposed.The milling process model and BP neural network control model are established respectively,in view of the problem that the constant force control ball end milling cutter cutting is difficult to implement due to the changes of multiple indicators,the autonomous height fitting characteristics of multi-layer BP neural network is used,combined with the constant force control multi-time cutting force and related servo input,the self-adaptive adjustment of multiple indicators is achieved to complete the constant force control.Among them,the index includes cutting force,feed rate,feed rate,back tool feed.In order to further explore the influence of noise on the robustness of the control algorithm,the square wave noise is used to carry out interference experiment on the established control algorithm,and the anti-interference ability of the control algorithm is discussed.The results show that by using the proposed control method,the constant force of ball end milling cutter can be tracked quickly and accurately,and the cutting force can be adjusted adaptively after the cutting depth is deepened;in the constant force control with added noise,the obtained simulation curve is consistent with that without added noise,which further proves the anti-interference ability of the proposed method.The proposed control method has good control performance and robustness.
作者
梁荣富
林朝光
李泽亮
王柱
王天雷
Liang Rongfu;Lin Chaoguang;Li Zeliang;Wang Zhu;Wang Tianlei(Jiangmen YIBINGDIN Machinery Co.Ltd.,Jiangmen,Guangdong 529030,China;Faculty of Intelligent Manufacturing,Wuyi University,Jiangmen,Guangdong 529020,China)
出处
《机电工程技术》
2023年第6期151-154,共4页
Mechanical & Electrical Engineering Technology
基金
国家自然科学基金项目(51505154)
广东高校优秀青年创新人才培养计划项目资助(2014KQNCX157)
江门市基础与理论科学研究类科技计划项目资助(2020JC01035)。
关键词
恒力切削
球头铣刀
BP神经网络
鲁棒性
constant force cutting
ball end milling cutter
BP neural network
robustness
作者简介
梁荣富(1968-),男,广东江门人,学士,工程师,研究领域为机械加工工艺及装备;李泽亮(1998-),男,广东韶关人,硕士研究生,研究领域为智能制造,深度学习图像处理技术;王柱(1976-),男,重庆人,博士,系统分析师,研究领域为智能制造,计算机系统与大数据分析;王天雷(1981-),男,湖南长沙人,博士,副教授,研究领域为智能制造,控制技术,计算方法。
