火电机组主蒸汽温度优化系统控制策略的研究

发布时间：2018-03-17 03:26

本文选题：主汽温控制系统　切入点：模糊自整定PID　出处：《华北电力大学》2016年硕士论文　论文类型：学位论文

【摘要】：能源是现代化的基础和动力。我国能源资源约束日益加剧,生态环境问题突出,调整结构、提高能效和保障能源安全的压力进一步加大,能源发展面临一系列新问题新挑战。随着我国国内经济形势进入“新常态”,经济发展对电力行业提出了更加严格的要求,在保证电力生产稳步推进的过程中,新建燃煤发电机组供电煤耗要低于每千瓦时300克标准煤,同时污染物排放接近燃气机组排放水平。燃煤机组供电煤耗的减少势必对电厂运行中的每个环节都提高要求,其中主蒸汽汽温的调节是保证机组设备安全生产、电力负荷稳定的基础。本文主要针对火电厂主蒸汽汽温控制策略进行研究,拟通过外挂控制器的方式在不同的负荷要求下,保证主蒸汽汽温的稳定,优化主蒸汽汽温的动态响应速度。本文通过虚拟仪器技术平台LabVIEW设计了主蒸汽汽温的外挂控制器,使用MODBUS TCP/IP协议,配合新华分散控制系统XDPS6.0,完成了外挂控制器的设计,并在心跳包逻辑的支持下,实现了外挂控制器的自动切换,既能保证主蒸汽汽温的稳定,又能在外挂控制器失效或者有手动操作需求的时候切换回传统PID控制器。外挂控制器的设计有两种思路:一种是利用模糊理论,在传统PID控制器的基础上,添加PID参数自整定功能,形成模糊自整定PID控制器;另一种是利用人工智能发展的结晶——神经网络技术,通过神经网络技术和模糊理论的结合,形成深度整合的模糊神经网络控制器。同时,针对传统PID控制器参数优化问题,使用帝国竞争算法予以寻优解决。最后,将两种外挂控制器和传统PID控制器对比,通过仿真实验证明了所提外挂控制器的实用性和先进性,并在对比的过程中得出不同控制器在不同场合具有各自的优点。本文的主要工作在于:1.在考虑被控对象约束的情况下,使用帝国竞争算法优化传统PID控制器参数;2.使用虚拟仪器技术实现外挂控制器的设计;3.完成心跳包逻辑设计,使外挂控制器更加实用,满足电厂安全性能要求;4.通过模糊自整定PID和模糊神经网络两种外挂控制器的设计,得出以下结论:模糊自整定PID控制策略的动态调节能力和鲁棒性较强,能够适应复杂多变的现场环境,而模糊神经网络控制策略适用于设备模型较为稳定、控制精度要求较高的场景。
[Abstract]:Energy is the foundation and motive force of modernization. The constraints of energy and resources in China are becoming increasingly severe, the ecological and environmental problems are prominent, and the pressure to adjust the structure, improve energy efficiency and ensure energy security is further increased. Energy development is facing a series of new problems and new challenges. As China's domestic economic situation enters the "new normal", economic development puts forward more stringent requirements for the power industry, and in the process of ensuring the steady progress of power production, The power consumption of new coal-fired generating units is lower than 300 grams of standard coal per kilowatt-hour, and the emission of pollutants is close to the emission level of gas-fired units. The reduction of coal consumption for coal-fired units will inevitably raise the requirements for every link in the operation of power plants. The regulation of main steam temperature is the basis to ensure the safe production of unit equipment and the stable power load. This paper mainly focuses on the main steam temperature control strategy of thermal power plant, and proposes to use external controller under different load requirements. To ensure the stability of the main steam temperature and optimize the dynamic response speed of the main steam temperature, the external controller of the main steam temperature is designed by the virtual instrument technology platform LabVIEW, and the MODBUS TCP/IP protocol is used. With XDPS6.0, the design of external controller is completed, and with the support of heartbeat packet logic, the automatic switch of external controller is realized, which can guarantee the stability of steam temperature of main steam. It can also switch back to the traditional PID controller when the external controller fails or has manual operation requirements. There are two ideas in the design of the external controller: one is to use fuzzy theory on the basis of the traditional PID controller. The fuzzy self-tuning PID controller is formed by adding the function of PID parameter self-tuning, the other is the combination of neural network technology and fuzzy theory through the use of crystal -neural network technology developed by artificial intelligence. The fuzzy neural network controller with deep integration is formed. At the same time, aiming at the parameter optimization problem of the traditional PID controller, the imperial competition algorithm is used to solve the problem. Finally, the two kinds of external controller and the traditional PID controller are compared. The practicability and advanced nature of the external controller are proved by simulation experiments, and in the process of comparison, it is concluded that different controllers have their own advantages in different situations. The main work of this paper is: 1. Under the condition of considering the constraint of the controlled object, the main work of this paper is: 1. Using Imperial Competition algorithm to optimize the parameters of traditional PID Controller 2.Using virtual instrument technology to realize the design of external controller. 3. To complete the logic design of heartbeat packet, make the external controller more practical, According to the design of fuzzy self-tuning PID and fuzzy neural network external controller, the following conclusions are drawn: the dynamic regulation ability and robustness of fuzzy self-tuning PID control strategy are strong. It can adapt to the complex and changeable field environment, while the fuzzy neural network control strategy is suitable for the scene where the equipment model is more stable and the control precision is higher.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP273;TM621

【参考文献】