摘要
桥梁温度是影响桥梁质量的重要因素,为找出桥梁温度的估算模型,以BP神经网络模型为基础,采用优化的麻雀算法对BP模型进行改进,构建了ISSABP模型,并将该模型计算结果与不同传统机器学习模型进行了比较,分析了模型精度,结果表明:桥梁温度在年内的变化趋势存在明显的规律性,ISSABP模型在桥梁温度模拟中的误差最低、一致性指标最高、可移植性最强,可推荐用于估算桥梁温度。
Bridge temperature is an important factor affecting bridge quality. In order to find out the estimation model of bridge temperature, based on BP neural network model, we used the optimized sparrow algorithm to improve the BP model, and constructed the ISSABP model. The calculation results of the proposed model were compared with those of traditional machine learning models, and the model accuracy was analyzed. The results showed that: The change trend of bridge temperature in the year had the obvious regularity. The ISSABP model had the lowest error, the highest consistency index and the strongest portability in bridge temperature simulation, which can be recommended for estimating bridge temperature.
作者
王微
Wang Wei(Beijing Municipal Bridge Maintenance Management Group Co.,Ltd.,Beijing 100097,China)
出处
《科学技术创新》
2022年第34期157-160,共4页
Scientific and Technological Innovation
关键词
桥梁温度
BP模型
麻雀算法
可移植性
bridge temperature
BP neural network model
sparrow algorithm
portability
