转炉煤气回收系统的PID神经网络控制策略研究  

Research on PID Neural Network Control Strategy in Converter Gas Recovery System

作　　者：郭许林[1] 章家岩[1] 吴宇平[1] 冯旭刚[1] 江涛[2] 

机构地区：[1]安徽工业大学电气信息学院,安徽马鞍山243032 [2]安徽工业大学建筑工程学院,安徽马鞍山243032

出　　处：《安徽工业大学学报（自然科学版）》2013年第3期328-332,共5页Journal of Anhui University of Technology（Natural Science）

基　　金：安徽省教育厅自然科学基金重点项目(KJ2013A054)

摘　　要：针对转炉炉口差压控制系统非线性、纯滞后和大干扰的特点,通过分析转炉煤气回收过程工艺及炉口差压对煤气回收效果的影响,提出将传统PID与神经网络控制策略相结合的炉口差压控制策略,并在对被控对象的辨识模型进行仿真分析的基础上,将控制策略投入工业应用。结果表明,与传统的PID控制策略相比,此控制策略具有较强的抗干扰能力,可显著改善控制系统的动态性能,并达到较好的煤气回收和烟气减排效果。Aiming at the characteristics of nonlinear, pure delay, strong interference of converter mouth differential pressure control system, by means of analyzing converter gas recovery process mechanism and the influence of converter mouth differential pressure on gas recovery effect, PID neural network control strategy is proposed. On the basis of simulation analysis of control system object identification model, the control strategy is applied in industry. Results show that this control strategy has stronger ability of anti-interference than the traditional PID. The strategy significantly improves the dynamic performance of control system and achieves an ideal effect on gas recovery and reduction of gas emission.

关 键 词：煤气回收 炉口差压 PID神经网络 差压控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]