摘要
针对目前大坝变形预测精度不理想的问题,提出改进遗传算法和BP神经网络相结合的预测模型(IAGA-BP)。通过改进标准遗传算法的选择算子,增加种群多样性,自适应调整交叉和变异概率,对BP网络初始的权阈值进行深度寻优。通过选取实例数据,对BP、GA-BP和IAGA-BP模型进行大坝变形预测分析,分析结果表明,IAGA-BP模型能提高变形预测精度,具备良好的稳定性。
Aiming at the dissatisfactory accuracy of current prediction of dam deformation,an improved adaptive genetic algorithm was propound to optimize BP neural network（IAGA-BP）.The population diversity was increased by improving the selection operator of standard genetic algorithm while the crossover and mutation probability were adjusted adaptively,eventually optimizing the initial weights and thresholds.The models of BP,GA-BP and IAGA-BP were analyzed based on practical data.Results demonstrate that the IAGA-BP algorithm shows superior accuracy and stability compared with the two counterparts.
作者
邢尹
陈闯
刘立龙
程胜
苏永柠
周威
XING Yin1,2 , CHEN Chuang3 , LIU Li-long1,2, CHENG Sheng1,2 , SU Yong-ning1,2, ZHOU Wei1,2(1. College of Geomatics and Geoinforrnation, Guilin University of Technology, Guilin 541004, China; 2. Guangxi Key Laboratory of Spatial information and Geomatics, Guilin University of Technology, Guilin 541004, China;3. College of Electrical Engineering and Control Science, Nanjing Tech University, Nanjing 211816, Chin)
出处
《计算机工程与设计》
北大核心
2018年第8期2628-2631,2686,共5页
Computer Engineering and Design
基金
国家自然科学基金项目(41664002)
广西自然科学基金项目(2015GXNSFAA139230)
广西空间信息与测绘重点实验室基金项目(14-045-24-03
14-045-24-10)
关键词
变形监测
遗传算法
BP神经网络
大坝
优化
deformation prediction
genetic algorithm
BP neural network
dam
optimization
