摘要
混凝土拱坝振型与频率参数中可能存在部分噪声数据,影响裂缝损伤检测精度,同时,当拱坝训练样本较多时,极易出现“维数灾难”,影响裂缝损伤检测效率,为此提出一种基于概率神经网络的尾矿库大坝混凝土拱坝裂缝损伤检测方法。求解尾矿库大坝混凝土拱坝系统振型与频率参数,针对拱坝振型与频率参数中可能存在的噪声数据,引入概率神经网络,将参数作为概率神经网络模型的输入,利用概率神经网络良好的容错性提升参数处理效果。采用K-Means算法确定模型中模式层的神经元节点数量及位置,提升模型检测效率。为了消除数据各尺度差异,将传入数据做标准化处理,通过模式层非线性算子计算求解计算输入样本与训练集中每个样本间的相似度,利用Bayes决策判定各数据是否属于裂缝损伤类别,完成拱坝裂缝损伤检测。仿真结果表明,所提方法检测拱坝裂缝损伤检测误差小、效率高,说明该方法可行。
There may be some noise data in the vibration mode and frequency parameters of the concrete arch dam,which affects the detection accuracy of crack damage.At the same time,when there are many training samples of arch dams,it is very easy to have"dimension disaster",which affects the detection efficiency of crack damage.Therefore,a method of crack damage detection for a concrete arch dam of tailings dam based on a probabilistic neural network is proposed.The vibration mode and frequency parameters of the concrete arch dam system of the tailings dam are solved.Aiming at the noise data that may exist in the vibration mode and frequency parameters of the arch dam,the probabilistic neural network is introduced.The parameters are taken as the input of the probabilistic neural network model,and the good fault tolerance of the probabilistic neural network is used to improve the parameter processing effect.The K-Means algorithm is used to determine the number andlocation of neuron nodes in the model layer to improve the efficiency of model detection.In order to eliminate the difference of each scale of data,the incoming data is standardized,and the similarity between the input sample and each sample in the training set is calculated through the nonlinear operator of the model layer.The Bayes decision is used to determine whether each data belongs to the crack damage category,and complete the crack damage detection of the arch dam.The simulation results show that the proposed method has small error and high efficiency in detecting arch dam crack damage,which indicates that the method is feasible.
作者
时鹏程
李保珠
SHI Peng-cheng;LI Bao-zhu(College of land and Resources Engineering,Kunming University of Science And Technology,Kunming Yunnan 650031,China)
出处
《计算机仿真》
2025年第11期276-280,共5页
Computer Simulation
基金
国家自然科学基金委员会地区科学基金项目(42267063)。
关键词
尾矿库
混凝土拱坝
裂缝损伤
概率神经网络
Tailing storage
Concrete arch dam
Crack damage
Probabilistic neural network
