基于BP神经网络的生物质气化建模

Modeling of Biomass Gasification Based on BP Neural Network

生物质发电是生物质能利用的一种重要技术,生物质气化是生物质发电的核心环节。国内外很多学者为改进生物质的气化效率在多方面对其进行了建模与研究,通常采用动力学方法进行生物质气化建模。生物质气化过程复杂,动力学建模需要详尽的物性参数,而这些参数往往难以直接获得。针对机理建模的缺陷,选取小麦秸秆作为实验对象,记录气化反应的初始参数和结果,再用神经网络拟合小麦秸秆的气化反应过程,建立基于BP神经网络的生物质气化模型,并依据实验数据对模型进行仿真验证。结果表明,模型对小麦秸秆气化反应过程特性具有较好的模拟预测作用。

Biomass power generation is an important technology of biomass energy utilization and biomass gasification is the core link of biomass power generation.In order to improve the gasification efficiency of biomass,many scholars at home and abroad model and research it in many ways,and they usually model the biomass gasification by dynamic method.The physical and chemical processes of biomass gasification are complex and dynamic modeling requires detailed physical parameters but these parameter are often difficult to obtain directly.Aiming at the defects of mechanism modeling,wheat straw was selected as the experimental object.The initial parameters and results of the gasification reaction were recorded,and then the gasification reaction process of the wheat straw was fitted by a neural network,the biomass gasification model based on BP neural network was established,and the model was verified by simulation according to the experimental data.The results showed that the model had agood simulation and prediction effect on the process characteristics of wheat straw gasification reaction.