精密矫直行程－挠度复合关联模型及行程精确预测研究

This project proposes a novel method for the establishment of straightening stroke-deflection relational model and the prediction of straightening stroke.Firstly,from the stress-strain information, stroke, and curvature, a stroke-deflection original model by the study on the elastic-plastic deformation will be established. Using anisotropic yield criteria of multi-factor, a method to solve the elastic-plastic constitutive model and to extract the material properties will be developed. Secondly, using the edge measurement system which is independently developed, the deflection measurement and research is performed. With the straightening tests, a spring-back control matrix will be built. The matrix is used to obtain the combined relational analysis data of deflection, span, spring-back, and stroke. Then, a stroke-deflection original model modified by the spring-back control matrix, and a stroke -deflection relational model based on spring-back control are proposed.This research aims to find the model solution and to predict the straitening stroke with the predictive functional control method. In addition, by optimizing the multi-point and multi-step straightening control algorithm for straightening load and with the multi-point and multi-step straightening control strategy, the accurate prediction and control test will be done. The precision straightening theory for controlling small elastic-plastic deformation will be worked out and be proved. The innovated theory from this research has a great significance on the control of small deformation of mechanical parts and improving the accuracy and quality of high precision machinery. The results of this research that has independent intellectual property rights will easily solve the prediction of straightening stroke problem that is currently very hard to deal with because of the high straightening accuracy requirement. The results also developed a theoretical foundation and technical knowledge for making high precision equipment that used to be imported.

提出建立精密矫直行程－挠度复合关联模型及行程精确预测的新方法。首先研究弹塑性形变及挠度和曲率变化规律，建立行程－挠度原始模型,同时采用多系数的各向异性屈服准则，研究求解弹塑性本构关系模型和识别提取材料性能参数的方法；然后运用自主研发的边缘测量系统进行挠度测量研究，并通过矫直试验获取挠度、跨距、回弹与行程关联数据，分析其复合关联度，建立回弹控制矩阵。最后提出回弹控制矩阵修正行程－挠度原始模型，建立基于回弹控制的行程－挠度复合关联模型，并应用预测函数控制方法来进行模型求解和实现矫直行程精确预测。通过优化多点多步控制算法，以精密直线导轨为对象，进行行程预测与控制试验，验证回弹控制的精密矫直行程精确预测机理。产生的理论创新，对控制机械零件弹塑性小变形和提高其精度具有理论指导意义；产生的自主知识产权成果，将改变矫直行程精确预测难题制约矫直精度的现状，为改变精密矫直技术装备依赖进口奠定理论技术基础。

本项目围绕精密矫直技术及矫直行程精确预测问题展开研究，目的是提高精密矫直精度。首先在矫直载荷－挠度关系模型基础上，结合弹塑性形变分析应力应变及挠度和曲率变化规律，研究矫直行程－挠度复合关联模型的表征方法和建模原理，并通过控制回弹量完成矫直行程预测；为了明确矫直过程中金相组织和材料性能参数之间的关联关系，采用多系数的各向异性屈服准则，建立了弹塑性小变形的本构关系模型；针对材料性能参数的非线性变化规律及材料硬化现象，运用模式识别算法，引入梯度信息提高遗传算法的寻优能力，对材料性能参数进行在线识别提取研究；然后运用多点多步矫直加载试验方法获取跨距、挠度、回弹与行程关系数据，引入预测函数控制基本方法，优化多点多步加载矫直行程的控制算法，实现高精度的多点多步精密矫直行程预测；为了实现工件的多挠度在线测量及误差识别，基于FPGA搭建了测量数据的在线采集与处理系统，主要对测量响应及传输速度进行分析研究，建立工况现场主要噪声信号标定数学模型，实现挠度测量信号的高速采集与处理和试验验证；在前期矫直试验机研发的基础上，开展矫直机系统的伺服系统优化匹配，以直线导轨为实例对象，进行精密矫直试验，验证本项目所提出的矫直行程－挠度复合关联模型及行程预测方法，并建立相关的矫直工艺数据库。.本项目的研究形成了金属条材精密矫直的相关关键技术，首先围绕矫直过程中跨距－回弹－挠度－行程之间的复合关联关系，建立了矫直行程－挠度复合关联度的表征方法和理论模型；为了实现精确的行程预测，考虑材料在循环载荷作用下的力学特性变化及存在的各向异性状态，展开对多系数的各向异性屈服准则和连续性介质的应力应变算法的研究，实现了弹塑性小变形的本构关系模型求解，并建立了相应的材料性能提取方法；有效结合矫直行程预测控制的基本理论方法和多点多步矫直综合试验，完成了多点多步加载矫直行程的控制算法的优化及实际应用，为相关矫直设备的设计制造提供了理论基础。.本项目的实施对矫直行程－挠度复合关联模型及其行程预测产生理论创新，对控制机械零件弹塑性小变形和提高其精度具有理论指导意义；提出的精密矫直行程预测新方法，将改变精密矫直行程精确预测难题制约精密矫直精度的现状，产生具有自主知识产权的研究成果，为精密矫直装备提供理论依据和技术支持，为改变我国精密矫直技术装备依赖进口并促进其高精度高精密化进程奠定理论基础和技术储备。