摘要:目前,我国现有很多工业和民用中小型锅炉,其中大部分自动控制水平很低,存在燃烧不彻底、排烟氧含量偏高、排烟热损失大等问题,直接造成了锅炉热效率低下和大量能源的浪费,同时也给环境带来很大的污染,所以实现锅炉的计算机控制具有重要的意义。

　　　首先，分析了热电厂锅炉控制的难点，以及采用常规控制方法难以获得良好控制效果的原因。其次，分析了锅炉燃烧过程机理及控制目标，确定锅炉燃烧系统是一个多变量的系统，其输入量为燃料量、送风量和引风量，输出量为主蒸汽压力、烟气含氧量和炉膛负压。

　　　论文根据某热电厂生产的实际台账数据，重点建立了锅炉燃烧系统多变量的神经元网络模型并采用模糊控制技术进行控制研究，即在总结操作人员成功经验的基础上，确定其模糊控制规则表并采用Matlab软件对系统进行仿真，仿真研究结果表明其控制策略具有较好的控制效果。

关键词：锅炉；燃烧系统；神经元网络；模糊控制；仿真

Abstract:At present, our country's existing industrial and civil many small and medium-sized boiler, most of them to be automatic control level is low, there is not complete combustion, smoke oxygen content is on the high side, the smoke heat loss of the problems, as a direct result of the thermal efficiency is low and boiler a waste of energy, but also to the environment a lot of pollution, so realize the computer control of the boiler is of great significance.

　　　Firstly, analyze the difficulty of boiler-control in thermal power plants, and the reason why adopting conventional methods couldn’t obtain good control results. Then, analyze the mechanism and the target of combustion process, confirmed that the combustion system of boiler was a multivariate system. The input quantity of this system was fuel quantity, supply air volume and air-input amount. While the output quantity was the pressure of main stream, the oxygen content of smoke, furnace negative pressure.

　　　The paper was made on the basis of the real data of a certain thermal power plant, and especially established the multivariate neural network model of boiler’s combustion system. In the mean time, adopt fuzzy control technology to study this system, In summary the operator based on the successful experience, identify the fuzzy control rule table and using Matlab software for the system simulation, the simulation results show that the control method has better control effect.

Key Words: Boiler； Combustion System； Neural network model； Fuzzy Control；Simulation