基于优化模糊神经网络的循环流化床锅炉自适应燃烧控制算法

摘要

There is some difficulty for the automatic and accurate control in the burning process of circulating fluidized bed boiler because of some complex features such as multivariate coupling and time-changing parameter,so an adaptive control er algorithm is proposed in this paper.The fuzzy BP neural network is combined with the fuzzy control and BP neural network by structural equivalence method,in which there are some defects including long-time for convergence so the ge-netic algorithm is introduced to optimize the weighting value.The adaptive and precise control for burning of circulating flu-idized bed boiler is realized by feed-forward compensation decoupling.%由于循环流化床锅炉燃烧过程的多变量耦合、参数时变等复杂特征导致其自动精确控制存在较大的难度，为此提出一种自适应控制器算法。将模糊控制与BP神经网络使用结构等价方法进行融合成模糊BP神经网络，并针对神经网络存在的收敛时间长等缺陷，引入遗传算法优化BP神经网络的权值，并通过前馈补偿解耦器实现对循环流化床锅炉燃烧过程的自适应精确控制。实验结果表明，该算法能够适应循环流化床锅炉的变参数工况，而且成功实现了燃烧过程中床温和主蒸汽压力的解耦合。