热轧带钢厚度智能控制方法研究.doc

摘　　要 随着国民经济的发展，钢铁的需求量急剧增加。钢铁工业已经成为国民经济发展的重要支柱。随着市场需求的不断发展，用户在对各种类型的板材需求急剧增长的同时，也对板材的各种材质精度提出了更高的要求，尤其在精度上需用厚度偏差仅为几个微米的热轧薄板。轧钢过程中的厚度控制是保证轧钢成品质量的重要因素。在实际生产过程中，由于液压AGC的模型是一个高阶的，非线性系统，在常规PID 控制器的控制过程中，存在超调量大，调节时间长，参数不易在线进行实时调整的缺点，难以满足实时性控制的要求。 神经元具有很强的信息综合、学习记忆和自学习、自适应能力，可以处理高阶的，非线性，时变的过程，将神经元与 PID 结合，应用于厚度的控制过程，给控制赋予智能，在线调整 PID 参数，使厚度控制系统具有较强的抗干扰能力、自适应能力和较好的鲁棒性。 本文运用MATLAB及其Simulink仿真软件对单神经元PID控制器和常规PID控制器进行研究比较。结果表明，应用单神经元自适应PID控制器来控制液压AGC要优于采用常规PID控制器下的液压AGC系统，单神经元自适应PID控制器可望在厚度控制中得到有效的应用。把先进的控制算法与成熟的厚控手段结合，进一步完善厚度控制系统，达到提高厚度控制精度的目的，从而为提高轧钢质量和产量做出新进展。 关键词： AGC，单神经元，PID，MATLAB仿真 Intelligent control of hot rolled strip of thickness Abstract With the development of national economy, steel demand increased dramatically. Iron and steel industry has become an important pillar of national economic development. With the growing market demand, the user needs in the rapid growth of various types of sheet metal, but also on the boards of the various materials put forward higher requirements for accuracy, especially in the accuracy required is only a few microns of thickness deviation The hot rolled sheet. The thickness of the rolling process control is to ensure that an important factor in the quality of rolling products. In the actual production process, because the hydraulic AGC is a high-end model, nonlinear system, conventional PID controller in the control process, there is large overshoot, long settling time, the parameter is not easy to adjust the shortcomings of online real-time Difficult to meet the requirements of real-time control. Neurons have a strong synthesis of the information, learning and memory and self-learning, adaptive capacity, can handle higher order, nonlinear, time-varying process, neurons and PID combination will be applied to the thickness of the control process, to the control Give smart, on-line adjustment of PID parameters, the thickness control system has strong anti-interference ability, adaptive ability, and better robustness. In this paper,