论文部分内容阅读
空气质量关系着人们的身体健康,因此研究实时、高效的空气质量预报系统,不仅能为公众出行提供指导,还能指导职能部门防控重污染天气并提供相应技术支持。近几年,国内外对预报理论及方法的研究主要集中在BP神经网络预报[1]。Deden Supriyatman[2]应用传统BP神经网络预报输气管道腐蚀速率,N.Haghdadi等人[3]应用改进BP网络预报半固态A356铝合金的热变形行为,对学习步长加入动量项,改善了收敛慢的问题,但
Air quality has a bearing on people’s health. Therefore, the study of real-time and efficient air quality forecasting system not only provides guidance for public travel, but also guides departments to prevent and control heavy pollutions and provide corresponding technical support. In recent years, the research on forecasting theory and method at home and abroad mainly focuses on BP neural network prediction [1]. Deden Supriyatman [2] applied traditional BP neural network to forecast the gas pipeline corrosion rate. N. Haghdadi et al. [3] applied the improved BP network to predict the thermal deformation behavior of semi-solid A356 aluminum alloy and added momentum term to the learning step to improve Slow convergence problem, however