论文部分内容阅读
基于概率神经网络原理,利用MATLAB神经网络函数,以27组生物质热解实验数据作为样本,建立了一个以着火点温度、峰值温度、最大失重速率、平均失重速率、燃烧特性指数为输入变量,以生物质类型为输出变量,能在生物质热解过程中进行识别的概率神经网络模型。模拟结果显示杆类、草类和炭类生物质在热解过程中均能正确地被识别。证明该模型具有较强的识别能力,可以对不同种类的生物质进行识别,从而为生物质精确燃烧提供指导。
Based on the principle of probabilistic neural network and using the function of MATLAB neural network, 27 groups of biomass pyrolysis experimental data were taken as samples to establish an ignition temperature, peak temperature, maximum weight loss rate, average weight loss rate and combustion characteristic index as input variables. A probabilistic neural network model that identifies the type of biomass as an output variable that can be identified during biomass pyrolysis. The simulation results show that the biomass of rods, grasses and charcoal can be correctly identified during pyrolysis. It is proved that this model has strong recognition ability and can identify different kinds of biomass so as to provide guidance for accurate combustion of biomass.