节点文献

分散控制系统(DCS)MAX1000+plus的PID参数自整定软件包开发研究

Developing and Researching PID Parameters Auto-tuning Software Package in Distributed Control System MAX1000+plus

【作者】 王乐天

【导师】 徐天芳;

【作者基本信息】 重庆大学 , 热能工程, 2003, 硕士

【摘要】 在控制工程实践中,PID控制是应用最为广泛的一种控制策略。PID参数的最佳整定对确保控制系统的控制品质起着至关重要的作用,只有整定出与被控对象动态特性相匹配的PID参数,才能满足控制系统的要求。然而,至今控制系统的PID参数整定工作无论是在国内还是在国外的控制工程实践中,主要采用的还是人工整定法。这使得PID参数的整定工作不仅费时,而且难以保证控制系统性能最佳,尤其是对于被控对象特性随工况发生变化的系统,PID参数人工整定法更不能满足系统的要求。因此,对PID参数自整定的研究就提到了议事日程。随着控制理论的飞速发展和先进控制设备DCS的采用,为解决PID参数自整定的研究提供了契机。 本文研究在分散控制系统(DCS)MAX1000+plus的上位机上开发PID参数的自整定软件包,根据被控对象的不同特性,采用了三种PID参数自整定策略,即自动整定技术(automatic tuning)、增益调度技术(gain scheduling)和PID参数自适应技术(PID adaptive control);然后根据被控对象的差异,采用了基于阶跃响应的过渡过程响应法和基于继电环的频率响应法两种可供选择的方案获取被控对象动态特性;再对多种PID参数的自整定算法进行比较分析后,借鉴了配置系统闭坏主导极点的基本思想,并把系统的最大灵敏度作为主要性能参考指标之一,作为PID参数的自整定算法;同时,针对复杂控制系统中应用最为广泛的串级控制系统与前馈控制系统,提出了与其相适应的PID参数自整定策略;在此基础上,利用Visual C++在MAX1000+plus的上位机上开发了PID参数的自整定软件包,并应用工程软件MATLAB建立被控对象模型,实现了与该DCS中虚拟DPU的连接,开发了PID控制仿真平台,对开发的PID参数自整定软件包进行了仿真试验。 仿真试验验证了该PID参数的自整定算法正确,具有一定的先进性和实用性,充分考虑了工业控制中各种被控对象的特点及差异,在实践上易于工程实际应用,能满足控制系统的要求。采用该算法开发的PID参数自整定软件包设计合理、界面良好、易于操作、功能完善、应用范围广,不仅适合于单回路控制系统的整定,而且也适合于串级控制和前馈控制复杂系统的整定。本文的研究工作使PID参数自整定算法在MAX1000+plus上予以实现,进一步发挥了分散控制系统(DCS)的功能。

【Abstract】 Proportional-Integral-Derivative (PID) is one of the most widely used control strategies in the process control engineering practice. The best tuning parameters of PID has a crucial effect on control performance. The desired requirements on control system can be met as long as the PID parameters accommodated with the process dynamics are obtained. However, the artificial methods for tuning PID parameters are so far applied commonly regardless of in the domestic or overseas process control engineering practice, which cause the PID tuning job extremely time-consuming and can not assure of the best control system performances consequently, particularly for these time variant processes. Hence, the research on automatic tuning and adaptation technologies for PID control is on the agenda. Furthermore, the rapid development of control theory and the application of advanced control equipment DCS provide an opportunity for PID parameters tuning.This paper researches the developing of PID parameters auto-tuning software package on distributed control system MAX1000+plus’s workstation, and adopts three strategies for tuning PID parameters in terms of the process characteristics, namely automatic tuning, gain scheduling and PID adaptive control. Then, two alternative schemes can be used to obtain the process dynamics. One is transient response method based on step test, the other is frequency response method based on relay feedback. The proposed PID parameters auto-tuning algorithm, which uses the idea of placement system’s dominant poles for reference and regards the maximum sensitivity as one of performance criteria, is concluded from the analyses of several algorithms. Meanwhile, the auto-tuning strategies focused on cascade and feedforward control systems are put forward as well. On the foundation of above, PID parameters auto-tuning software package on MAX1000+plus’s workstation is developed with Visual C++. Finally, the PID control simulation platform, which is composed of the process model established by MATLAB and model’s connection with virtual DPU, is developed for this software.The simulation results show that this tuning algorithm that has taken the processes characteristics differences into consideration is correct, advanced and practical in engineering application and meets the requirements on the control system. The software adopted this auto-tuning algorithm is rationally designed; its human machine interface is friendly; its functionality is versatile; and its application scope is broad, not only for single-loop control system, but for cascade and feedforward control systems. The research work in this paper realized PID auto-tuning algorithm on MAX1000+plus’s workstation and further developed DCS’s function.

【关键词】 分散控制系统PID参数自整定软件包
【Key words】 DCSPID parametersauto-tuningsoftware package
  • 【网络出版投稿人】 重庆大学
  • 【网络出版年期】2003年 04期
  • 【分类号】TK323
  • 【下载频次】238
节点文献中: 

本文链接的文献网络图示:

本文的引文网络