摘要
在探讨多源数据融合技术和引进遗传算法(GA)的改进粒子群(PSO)优化BP神经网络(GA-PSO-BP)模型在滑坡监测和预测领域的应用及其效果的基础上,以福建省安溪县西坪镇滑坡监测数据为例,研究验证集成多种数据源及采用粒子群算法优化BP神经网络的有效性。结果表明,GA-PSO-BP模型能显著提高滑坡监测的精确度与可靠性,有效解决了BP神经网络易陷入局部最优解和对训练数据需求高的问题,预测滑坡位移的平均绝对误差(MAE)和均方根误差(RMSE)均低于传统方法,展现出较高的预测性能。在处理具有高度相关性和冗余性的多源数据时,集中式和分布式数据融合方法的有效性为滑坡预警系统提供了新的策略和方法。
Based on the exploration of multi-source data fusion technology and the application of the improved Genetic Algorithm(GA)-Particle Swarm Optimization(PSO)-BP Neural Network(GA-PSO-BP)model in the field of landslide monitoring and prediction,this study uses landslide monitoring data from Xiping Town,Anxi County,Fujian Province as an example to verify the effectiveness of integrating multiple data sources and optimizing BP neural networks with particle swarm algorithm.The results show that the GA-PSO-BP model can significantly improve the accuracy and reliability of landslide monitoring,effectively solve the problems of BP neural networks being prone to local optima and high demands for training data.The model s prediction of landslide displacement demonstrates lower Mean Absolute Error(MAE)and Root Mean Square Error(RMSE)compared to traditional methods,indicating superior predictive performance.In processing multi-source data with high correlation and redundancy,the effectiveness of centralized and distributed data fusion methods offers new strategies and approaches for landslide early warning systems.
作者
蔡伟佳
聂闻
霍蔚然
CAI Weijia;NIE Wen;HUO Weiran(College of Advanced Manufacturing Engineering,Fuzhou University,Quanzhou 362251,Fujian,China;Quanzhou Equipment Manufacturing Research Center,Haixi Institute,Chinese Academy of Sciences,Quanzhou 362200,Fujian,China;College of Electrical Engineering and Automation,Fuzhou University,Fuzhou 350108,Fujian,China)
出处
《水力发电》
CAS
2024年第8期16-21,共6页
Water Power
基金
国家自然科学基金资助项目(41072232)
福建省科学院科学技术合作计划(2022T3051)。
关键词
滑坡监测
多源数据融合
BP神经网络
粒子群优化算法
遗传算法
landslide monitoring
multi-source data fusion
BP Neural Network
Particle Swarm Optimization Algorithm
Genetic Algorithm
