基于模糊神经网络的锅炉床温控制策略研究  

Research on Boiler Bed Temperature Control Strategy Based on Fuzzy Neural Network

作　　者：胡兴 韩磊 HU Xing;HAN Lei(Jinkong Power Tongda Thermal Power Shanxi Co.,Ltd.,Datong 037001,China;School of Computer&Information Technology,Shanxi University,Taiyuan 030006,China)

机构地区：[1]晋控电力同达热电山西有限公司,山西大同037001 [2]山西大学计算机与信息技术学院,山西太原030006

出　　处：《电工技术》2023年第6期3-6,共4页Electric Engineering

摘　　要：针对传统PID控制策略对非线性、时变、多变量耦合、大迟延和惯性的循环流化床床温控制品质不佳,提出运用模糊控制及神经网络相结合的方法对锅炉燃烧系统中的床温被调量进行控制。由模糊控制与神经网络深度融合形成的模糊神经网络控制方式结合了模糊推理与神经网络的并行处理、自学习的优点,网络通过训练可以实现控制规则的自动辨识与隶属度函数的校正,实现“智能化”的调节要求。在MATLAB仿真软件中搭建仿真模块,仿真结果表明该控制策略较传统PID控制平稳调节性能更好,达到稳态的时间更短。The temperature control of circulating fluidized bed bed has the characteristics of nonlinearity,time-vary,multivariable coupling,large delay and inertia,and the traditional PID control strategy has a poor control effect.A combination of fuzzy control and neural network is proposed to control the bed temperature in the boiler combustion system.The fuzzy neural network control mode formed by the deep fusion of fuzzy control and neural network combines the advantages of parallel processing and self-learning of fuzzy reasoning and neural network.Through training,the network can realize the automatic identification of control rules and the correction of membership function,and realize the"intelligent"adjustment requirements.Simulation modules are built in MATLAB simulation software.The simulation results show that the control strategy has better smooth adjustment performance than the traditional PID control and the time to reach steady state is shorter.

关 键 词：模糊神经网络 循环流化床锅炉 床温控制 智能化调节 

分 类 号：TK321[动力工程及工程热物理—热能工程]