《精品》-单神经元PID控制在热连轧监控AGC上的应用.doc

单神经元PID控制在热连轧监控AGC上的应用 摘要 热轧带钢厚度精度一直是提高产品质量的主要目标。产生厚差的原因主要有带钢头尾部温差、加热炉板坯生成水印以及温度随机波动。目前消除厚差的主要办法是采用自动厚度控制系统（AGC）(automatic gauge control)，因此,自动厚度控制（AGC）一直是热轧带钢自动化首先实现的功能。AGC系统的主要任务是对带钢全长进行厚度控制以保证带钢的厚度精度及其百分比。由于实际轧制过程的复杂性、控制对象的非线性、时变性、单纯的AGC控制系统都不能取得较好的控制效果。因此在很多热轧现场上都采用综合AGC控制策略，将先进的控制技术引入轧制生产中，无疑为传统AGC控制技术开辟了更广阔的天地。本文以监控AGC模型为研究对象，给出了基于单神经元网络控制算法的硬件实现方法和具体步骤，主要以PID控制器为基础,辅助以单神经元网络的自学习、自适应能力,将单神经元网络和PID控制融合为一体，提出了一种单神经元网络PID控制器（SNPC）应用于监控AGC上，进一步提高了AGC的厚控精度,有效的改善了系统的滞后性和动态品质，并具有良好的自适应能力。 关键字：热连轧；单神经元网络；PID控制；监控AGC  Based on the neural network of PID control in the hot strip mill AGC control of the Applied Research Abstract Gauge precision of hot rolling strip is always primary goal of improving product quality. The main reason for producing gauge difference is the gauge difference of strip’s first and end, slab reheating furnace and the temperature generated watermark random fluctuations. Now, the main way of eliminate gauge difference is adopt automatic gauge control system (AGC) (automatic gauge control), therefore, the automatic gauge control (AGC) has been hot strip first to achieve automation functions. The main assignment of AGC is to control the whole hot rolling strip’s length to ensure it’s gauge precision and percentage. The result of Simple AGC system can’t obtain preferable effect due to the complex of rolling process and non-linearity & time-variety of controlled object. Synthetically AGC and advanced process control are applied in many hot strip mill,without doubt, it has a promising and extensive future in application than conventional AGC system. This article’s study object is monitor AGC ,include the hardware control method and the specific steps based on the single neural network, the basis is PID control system and the single neural network’s self-learning capacity ,self-adapting capacity is assistance, combine the single neural network PID controllable equipment, apply it to monitor AGC .It improve the gauge precision of AGC ,the lagging and d