粉状炸药生产线改性塔压力控制技术研究

【作者】 李钊勋；

【导师】 王随平；

【作者基本信息】 中南大学 ， 控制科学与工程， 2009， 硕士

【摘要】 改性塔是改性粉状硝铵炸药生产线中对硝酸铵进行改性的装置,作为粉状炸药生产过程中的一个关键设备,其压力控制的好坏与否直接关系到炸药质量的高低,对整个系统的安全性具有极大的影响。因此,改性塔压力自动控制系统的研究与设计对于实现粉状炸药生产过程的全自动化、提高产品质量、降低劳动强度具有非常重要的意义。本文系统介绍了粉状炸药的生产工艺,分析了改性塔压力的产生及其特点。在此基础上对改性塔压力系统进行机理建模得到模型结构,并通过测试建模法辨识出模型参数,得到了改性塔压力系统的近似模型。改性塔压力系统具有时变性和滞后性的特点。本文分析了滞后环节的产生及其对系统性能的影响,详细介绍了Smith预估控制原理,着重研究了单神经元自适应PID控制器的设计以及对算法的改进。Smith预估控制是基于模型的补偿控制,但它对参数变化较为敏感。基于此,本文提出了一种单神经元自适应Smith预估补偿控制策略,将单神经元自适应PID作为Smith预估控制的主控制器,而将Smith预估器作为补偿器。利用MAILAB中的Simulink工具对控制器进行仿真,仿真结果表明,该控制方案具有良好的控制品质和鲁棒性。本文最后介绍了改性塔压力控制系统的结构,同时进行了控制系统的软件设计。现场运行情况表明,所采用的控制器稳定可靠,实现了粉状炸药的连续化生产,提高了产品的产量和质量。更多还原

【Abstract】 Reaction tower, which is one of the key units in the process of powdery explosive production, is used to modify the character of ammonium nitrate in the production line of powdery explosive. The pressure stability of the reaction tower influences not only the performance of the dynamite but also the security of its producing system. Therefore, the research and design for the automatic pressure control system of the reaction tower are of great significance for realizing the automatic control in the manufacture of powdery explosive, improving the quantity of product and decreasing working intensity.Based on the analyses for the production process of the powdery explosive and the characteristic of the reaction tower-pressure system, a model structure of reaction tower-pressure control system is established firstly, and then the system parameters are identified by using classical step response method. An approximate system model is deduced at last.The main characteristics of the reaction tower-pressure control system are time-delay and time-varying. The influence which time-delay loop brings to the system performance and the principle of the Smith predictor are introduced particularly in this paper. The theory of single neuron adaptive PID controller is researched, mainly in its algorithm and the algorithm modification. Smith predictor can effectively remove the time-delay from the closed-loop characteristic equation, but it is sensitive to the change of the parameters. In order to solve this problem, a strategy of single neuron self-adapting compensation control via Smith predictor is put forward. In this strategy, the single neuron PID is used as the main controller, while Smith predictor is used as the compensator. The intelligent controller is simulated with Simulink in Matlab and the results show that the proposed scheme has a good control character and strong robustness.Finally the structure of the system is introduced and the software system is designed. The running results indicate the control system is stable and reliable. The continuous production is successfully realized. The quantity and quality of product are also improved.更多还原