摘要
高精度几何量计量误差补偿面临多源耦合、系统复杂等技术难题。通过构建温度场、振动场、湿度场多场耦合误差模型,揭示了环境因素对测量精度的影响规律。建立随机—系统混合误差数学模型,实现了误差特性的准确表征。提出基于深度学习和模糊神经网络的智能补偿算法,设计了在线迭代优化方法。实验表明,本文所提补偿方法使测量精度提升45%,动态响应时间缩短30%。工程应用验证了该补偿系统具有较强环境适应能力,测量重复性优于0.15μm,为高精度几何量计量提供了可靠技术支撑。
High-precision geometric quantity metrology error compensation faces technical challenges such as multi-source coupling and system complexity.By constructing a multi-field coupling error model that includes temperature,vibration,and humidity fields,the influence patterns of environmental factors on measurement accuracy are revealed.A mathematical model for random-system hybrid errors is established to achieve accurate characterization of error characteristics.An intelligent compensation algorithm based on deep learning and fuzzy neural networks is proposed,and an online iterative optimization method is designed.Experiments show that the compensation method proposed in this article improves measurement accuracy by 45%and reduces dynamic response time by 30%.Engineering applications have verified that the compensation system has strong environmental adaptability and measurement repeatability better than 0.15μm,providing reliable technical support for high-precision geometric quantity metrology.
作者
雷毅
王露
贺闫平
王阿凡
Lei Yi;Wang Lu;He Yanping;Wang Afan(AECC Aviation Power Co.,Ltd.,Xi'an,Shaanxi,China,710021)
出处
《仪器仪表用户》
2025年第5期36-38,共3页
Instrumentation
关键词
几何量计量
多场耦合
误差补偿
深度学习
在线优化
geometric quantity metrology
multi-field coupling
error compensation
deep learning
online optimization
