【作者】 郭斌；

【导师】 孟令启；

【作者基本信息】 郑州大学 ， 机械制造及其自动化， 2011， 硕士

【摘要】 中厚板轧件的厚度精度是轧件重要的质量指标之一,厚度控制的实质是对轧机的辊缝进行控制。本文分析了影响轧件出口厚度的因素及轧机厚度的控制原理,利用GRNN神经网络建立了4200轧机钢板厚度的预测模型,结果表明GRNN神经网络模型能较好地预测轧件厚度的变化情况,其相对误差很小。相对于BP神经网络和Elman神经网络模型,GRNN神经网络模型在预测精度方面有一定程度的提高,实现了与实测结果的高度拟合。人工神经网络为解决厚度预测提供了一种新的解决途径,特别是在轧钢过程中有广泛的应用前景,这在理论以及实践上都将具有重要的意义。轧机厚度控制系统是轧机最重要的部分之一,本文介绍了常规PID控制算法的理论基础,分析了模糊PID控制较之于常规PID控制的优势,并详细叙述了模糊PID控制器的设计过程。利用MATLAB模糊逻辑工具箱,建立了以辊缝位置为被控对象的参数自调节模糊PID厚度自动控制系统的模型,对轧机自动控制系统进行模糊PID控制。在MATLAB/Simulink环境下,分别建立了单位阶跃响应的模糊PID控制器的系统模型和单位脉冲响应的模糊PID控制器的系统模型。通过比较常规PID控制器和模糊PID控制器的响应结果可知,模糊PID控制器控制下的液压系统的超调量比常规PID控制器小,这说明模糊PID控制器的性能优于常规PID控制器,这为解决厚度自动控制提供一种有效的途径,为轧机厚度控制技术的发展提供了很好的前景,同时也为现代工业自动控制问题提供了新的解决方案。更多还原

【Abstract】 The thickness of plate rolling precision is an important indicator of the quality indicators, thickness control, in essence, is rolling mill roll gap control. In this paper, the exports of their impact on the thickness of rolled pieces of various factors and several mill thickness control principle were analyzed. And using GRNN neural network of 4200 mill thickness of predictive models of the thickness forecast. The results show that GRNN neural network model can predict the thickness well, and it has very small relative errors. Through compared with BP network and Elman network, GRNN neural network has improved to some extent in prediction accuracy, achieved a high degree of fit with the measured results. Artificial neural network provides a new solution to solve the AGC problem, especially in the rolling process has broad application prospects, and there will be significance in both theory and practice.Automatic gauge control system is one of the most important parts of the steamrolling plant. This paper describes the conventional PID control algorithm theoretical basis, and analyzes the fuzzy PID control compared to conventional PID control advantages, and detailed description of the fuzzy PID controller design process. A roll seam location parameters of the controlled object the thickness of self-adjusting fuzzy PID control system model was established by MATLAB Fuzzy Logic Toolbox, and the system of plate rolling mill was controlled by fuzzy PID control system. The system of unit step response model of the fuzzy PID controller and the system of unit impulse response of the fuzzy PID controller were established in MATLAB/Simulink environment. By comparing the conventional PID control and fuzzy PID control of the response results, the fuzzy PID controller under the control of the hydraulic system than the conventional PID controller, the overshoot is small, which shows the performance of fuzzy PID controller is superior to conventional PID controller. It provides an effective means for the rolling mill thickness control technology and provides a very good prospect, and provides a new solution for the modern industrial automation at the same time.更多还原