面向燃气电动调压的PID神经网络控制系统设计与FPGA实现

发布时间：2019-01-26 20:22

【摘要】：开发利用天然气这一清洁能源,实施“西气东输”、“缅气工程”、“川气东送”、“陕气进京”等一系列重大输气工程,是实现我国能源结构调整、发展低碳经济的重大战略决策,这给燃气输配系统的建设与发展带来了新的机遇和挑战。作为燃气配送过程中的关键环节,燃气调压器的安全性和高效性越来越受到人们的重视。现有的燃气调压器通常采用传统的直接作用式或间接作用式的机械调压模式,这种机械调压方式调压精度低、稳定性差。由此,会导致燃气热值下降,造成能源的极大浪费;同时,产生大量一氧化碳,造成环境污染,甚至可能诱发供气安全等问题。因此,提高燃气调压器的调压性能,已成为燃气调压技术领域的前沿课题。本文在深入研究现有调压技术和控制理论的基础上,以需求量最大的低压燃气调压器为研究对象,提出并设计了一种新型的电动调压模式。探索了电动燃气调压测控问题的新理论和新方法,提出一种基于PID神经网络的燃气调压测控方案,并采用FPGA实现了PID神经网络的燃气调压实时测控。本文的研究工作,有望提升燃气调压器的技术水平,以提高燃气调压的实时性、准确性和可靠性,为燃气调压器的工程设计提供科学依据。论文的主要工作包括:研究了现有的燃气调压技术,建立了电动燃气调压的智能化测控方案;研究了工业中常用的PID控制技术;针对现有控制算法在电动燃气调压测控的具体应用问题,研究了人工神经网络算法,将人工神经网络与PID控制规则相融合,建立了基于PID神经网络的燃气调压控制算法;并用MATLAB对传统PID控制算法和PID神经网络算法进行仿真对比,结果表明PID神经网络比PID控制具有更加优越的性能;同时,用FPGA硬件平台进行了算法实现;最后以燃气调压器的气压控制系统为研究基础,对整个控制模块进行空载时序仿真和加入燃气调压器系统模型仿真,仿真结果表明:PID神经网络气压控制模块具有很高的控制精度和实时性,适用于解决大滞后、非线性调压系统的智能控制问题。最后对本课题所做的研究进行了总结,并对课题下一步的工作进行了展望。
[Abstract]:The development and utilization of natural gas as a clean energy source and the implementation of a series of major gas transmission projects, such as "West to East Gas Transmission", "Myanmar Gas Project", "Sichuan Gas to the East", and "Shaanxi Gas to Beijing", are a series of major gas transmission projects to realize the adjustment of China's energy structure. The important strategic decision of developing low-carbon economy brings new opportunities and challenges to the construction and development of gas transmission and distribution system. As a key link in gas distribution process, people pay more and more attention to the safety and efficiency of gas regulator. The traditional mechanical regulating mode of direct action or indirect action is usually used in the existing gas regulator, which has low precision and poor stability. As a result, the gas calorific value will decrease, resulting in a great waste of energy, at the same time, a large amount of carbon monoxide will be produced, resulting in environmental pollution, and even may lead to the safety of gas supply and so on. Therefore, improving the pressure regulation performance of gas regulator has become a frontier topic in the field of gas pressure regulation technology. On the basis of deep research on the existing voltage regulation technology and control theory, a new type of electric voltage regulating mode is proposed and designed in this paper, taking the low-pressure gas regulator with the greatest demand as the research object. In this paper, the new theory and method of electric gas pressure regulation and measurement and control are explored, and a new method of gas pressure regulation measurement and control based on PID neural network is put forward, and the real time measurement and control of gas pressure regulation based on PID neural network is realized by FPGA. The research work in this paper is expected to improve the technical level of gas regulator, in order to improve the real-time, accuracy and reliability of gas pressure regulation, and provide scientific basis for the engineering design of gas regulator. The main work of this paper is as follows: the existing gas pressure regulation technology has been studied, the intelligent measurement and control scheme of electric gas pressure regulation has been established, the PID control technology commonly used in industry has been studied. Aiming at the specific application of the existing control algorithms in the electric gas pressure regulation and measurement and control, the artificial neural network algorithm is studied, and the artificial neural network and PID control rules are combined to establish the gas pressure regulation control algorithm based on the PID neural network. The traditional PID control algorithm and PID neural network algorithm are simulated and compared with MATLAB. The results show that PID neural network has better performance than PID control, and the algorithm is implemented on FPGA hardware platform. Finally, based on the air pressure control system of the gas regulator, the simulation of the whole control module is carried out on the no-load timing simulation and the model simulation of the gas regulator system. The simulation results show that the PID neural network air pressure control module has high control precision and real-time performance, and it is suitable for solving the intelligent control problem of large lag and nonlinear voltage regulating system. Finally, the research of this topic is summarized, and the future work of the project is prospected.
【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TE97;TP273

【相似文献】