基于RBF神经网络的军粮城发电厂锅炉燃烧优化分析与改造

发布时间：2018-08-02 15:39

【摘要】：进入新世纪以来,为了应对能源紧张和环境恶化的双重挑战,我国加大了对火力发电企业节能减排的要求。发电企业也根据自身特点对硬件设备进行了最大限度的改良,但在复杂的运行条件下单靠设备改造已经不能满足新的要求。随着神经网络技术的发展,已经能够解决线性不可分的多层神经网络的学习问题。因此,对火电厂锅炉燃烧系统的运行调整进行神经网络优化就成为了进一步提高火电厂节能减排水平的很好选择。目前,国内一些专业院校开发了很多相关锅炉节能指导系统。但这些系统普遍都基于“DCS偏置量的修正”或者“经典控制理论”,大都数所针对的问题较为片面,不能综合考虑机组各系统之间的协调,不能兼顾机组效率的提高和环保排放要求,也不能给运行人员的操作提供及时的具体建议。本文所述的锅炉燃烧性能优化系统,是基于先进的RBF人工神经网络技术,结合国内发电企业的负荷及燃煤条件,进行有针对性的建模和简化设计而成的。具有训练速度快、实时性强、指导可靠性高等特点。可以在兼顾锅炉系统安全限制条件的前提下最大挖掘系统设计冗余,适应机组当前工况,综合提高燃烧系统整体效率,从而达到燃煤电厂燃烧调整的综合优化。
[Abstract]:Since entering the new century, in order to meet the dual challenges of energy stress and environmental deterioration, China has increased the energy saving and emission reduction requirements for thermal power enterprises. According to their own characteristics, the power generation enterprises also improve the hardware equipment to the maximum extent, but in the complex operation conditions, the order can not meet the new requirements by the equipment transformation. With the development of neural network technology, it has been possible to solve the learning problem of linear indivisible multilayer neural network. Therefore, neural network optimization of boiler combustion system in thermal power plant becomes a good choice to further improve the level of energy saving and emission reduction. At present, some professional colleges and universities have developed a lot of related boiler energy-saving guidance system. However, these systems are generally based on "DCS bias correction" or "classical control theory". Can not take into account the efficiency of the unit and environmental emission requirements, and can not provide operational personnel with timely specific advice. The boiler combustion performance optimization system described in this paper is based on the advanced RBF artificial neural network technology, combined with the load of domestic power generation enterprises and coal-burning conditions, targeted modeling and simplified design. Has the training speed fast, the real-time strength, the instruction reliability and so on characteristic. Under the premise of taking into account the safety limitation of boiler system, the maximum system design redundancy can be excavated, which adapts to the current working conditions of the unit and comprehensively improves the overall efficiency of the combustion system, so as to achieve the comprehensive optimization of the combustion adjustment in the coal-fired power plant.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM621.2;TP183

【相似文献】