以压力容器气体泄漏展开研究,提出了一种融合黄金正弦的减法平均优化器(subtraction-average-based optimizer with golden sine,GSABO)、优化变分模态分解(variational mode decomposition,VMD)和卷积神经网络(convolutional neural ne...以压力容器气体泄漏展开研究,提出了一种融合黄金正弦的减法平均优化器(subtraction-average-based optimizer with golden sine,GSABO)、优化变分模态分解(variational mode decomposition,VMD)和卷积神经网络(convolutional neural network,CNN)与支持向量机(support vector machine,SVM)联合分类检测的方法。首先,引入了融合黄金正弦的减法平均优化器对变分模态分解的参数模态个数K和惩罚参数α进行寻优,将最小包络熵为适应度函数得到最佳的K和惩罚参数α,计算最佳IMF分量的9种时域指标构建特征向量,输入CNN-SVM联合的分类方法进行特征提取并对气体泄漏情况进行识别。经实验分析,提出的引入融合黄金正弦的减法平均优化器优化后的VMD方法能够有效地自适应获取最优参数组,然后对压力容器气体泄漏声波信号进行特征提取,选取最优的特征组合输入CNNSVM联合分类检测,得到泄漏与否判别准确率高达99.16%,有助于对后续研究进一步开展。展开更多
准确的高铁沿线风速预测是铁路灾害预警系统的基础需求,为了提升应对和处理强风灾害致突发事件的能力,提出一种基于减法平均优化(subtraction average based optimizer,SABO)算法优化长短时记忆(long short-term memory,LSTM)神经网络...准确的高铁沿线风速预测是铁路灾害预警系统的基础需求,为了提升应对和处理强风灾害致突发事件的能力,提出一种基于减法平均优化(subtraction average based optimizer,SABO)算法优化长短时记忆(long short-term memory,LSTM)神经网络的高铁沿线短期风速预测方法。首先,针对风速非线性和非平稳特性,采用极小化极大(min-max,MM)方法对风速数据进行归一化处理;其次,采用SABO算法中的“-v”方法对LSTM模型的关键参数搜索寻优,并构建风速预测模型;最后,以中国宝兰高铁沿线风速采集点采集的实测风速数据为例,对模型进行有效性检验。实验结果表明:SABO算法的寻优效果更加良好,预测精度更高,所建模型的平均绝对误差(mean absolute error,MAE)、平均绝对百分比误差(mean absolute percentage error,MAPE)和均方根误差(route mean square error,RMSE)分别仅为11.96%、1.23%和16.47%,决定系数(r-square,R^(2))为0.995。与其他模型相比,通过SABO算法优化后的LSTM神经网络在短期风速预测上具有较好的拟合效果和更高的预测精度,可为高铁沿线大风预测预警提供一种新的方法和思路。展开更多
提出一种基于SABO-GRU-Attention(subtraction average based optimizer-gate recurrent unitattention)的锂电池SOC(state of charge)估计方法。采用基于平均减法优化算法自适应更新GRU神经网络的超参数,融合SE(squeeze and excitation...提出一种基于SABO-GRU-Attention(subtraction average based optimizer-gate recurrent unitattention)的锂电池SOC(state of charge)估计方法。采用基于平均减法优化算法自适应更新GRU神经网络的超参数,融合SE(squeeze and excitation)注意力机制自适应分配各通道权重,提高学习效率。对马里兰大学电池数据集进行预处理,输入电压、电流参数,进行锂电池充放电仿真实验,并搭建锂电池荷电状态实验平台进行储能锂电池充放电实验。结果表明,提出的SOC神经网络估计模型明显优于LSTM、GRU以及PSO-GRU等模型,具有较高的估计精度与应用价值。展开更多
Debris flows are one of the common natural hazards in mountainous areas.They often cause devastating damage to the lives and property of local people.The sabo dam construction along a debris flow valley is considered ...Debris flows are one of the common natural hazards in mountainous areas.They often cause devastating damage to the lives and property of local people.The sabo dam construction along a debris flow valley is considered to be a useful method for hazard mitigation.Previous work has concentrated on the different types of sabo dams such as close-type sabo dam,open-type sabo dam.However,little attention has been paid to the spillway structure of sabo dam.In the paper,a new type of spillway structure with lateral contraction was proposed.Debris flow patterns under four different spillway structures were investigated.The projection theory was employed to predict trajectory of debris flow out from the spillway and to estimate the incident angle and terminal velocity before it plunged into the scour hole behind the sabo dam.The results indicated that the estimated data were in good agreement with the experimental ones.The discrepancy between the estimated and experimental values of main parameters remained below 21.82%(relative error).Additionally,the effects of debris flow scales under different spillway structures were considered to study the scour law.Although the debris flow pattern and scour law behind the sabo dam under different operating conditions was analyzed in this paper,further study on the scour mechanism andthe maximum scour depth estimation based on scour theory is still required in the future.展开更多
The erosion shape and the law of development of debris flow sabo dam downstream is a weak part in the study on debris flow erosion. The shape and development of scour pit have an important effect on the stability and ...The erosion shape and the law of development of debris flow sabo dam downstream is a weak part in the study on debris flow erosion. The shape and development of scour pit have an important effect on the stability and safety of debris flow sabo dam, which determines the foundational depth of the dam and the design of protective measures downstream. Study on the scouring law of sabo dam downstream can evaluate the erosion range and reasonably arrange auxiliary protective engineering. Therefore, a series of flume experiments are carried out including different debris flow characteristics （density is varying from 1.5 t/m3 to 2.1 t/m~） and different gully longitudinal slopes. The result shows that the scour pit appears as an oval shape in a plane and deep in the middle while superficial at the ends in the longitudinal section, the position of the maximum depth point moves towards downstream with an increase of flume slope angle. The maximum depth of scour pit is mainly affected by the longitudinal slope of gully, density of debris flow, and the characteristics of gully composition （particle size and the viscosity of soil）. The result also indicates that the viscosity of soil will weaken the erosion extent. The interior slopes of scour pit are different between the upstream and the downstream, and the downstream slope is smaller than the upper one. For the viscous and non-viscous sands with the same distribution of gradation, the interior slope of non- viscous sand is smaller than the viscous sand.According to tbe regression analysis on the experimental data, the quantitative relationship between the interior slope of scour pit, slope of repose under water and the longitudinal slope of gully is established and it can be used to calculate the interior slope of scour pit. The results can provide the basis for the parameter design of the debris flow control engineering foundation.展开更多
In order to improve the accuracy of wind turbine fault diagnosis,a wind turbine fault diagnosis method based on Subtraction-Average-Based Optimizer(SABO)optimized Variational Mode Decomposition(VMD)and Kernel Extreme ...In order to improve the accuracy of wind turbine fault diagnosis,a wind turbine fault diagnosis method based on Subtraction-Average-Based Optimizer(SABO)optimized Variational Mode Decomposition(VMD)and Kernel Extreme Learning Machine(KELM)is proposed.Firstly,the SABO algorithm was used to optimize the VMD parameters and decompose the original signal to obtain the best modal components,and then the nine features were calculated to obtain the feature vectors.Secondly,the SABO algorithm was used to optimize the KELM parameters,and the training set and the test set were divided according to different proportions.The results were compared with the optimized model without SABO algorithm.The experimental results show that the fault diagnosis method of wind turbine based on SABO-VMD-KELM model can achieve fault diagnosis quickly and effectively,and has higher accuracy.展开更多
针对Kubernetes默认水平伸缩策略在高并发场景下因间歇时间而导致集群规模无法及时扩展,进而易引发集群性能下降甚至宕机的问题,提出了一种基于加权变分减法双向长短期记忆网络模型(INFO-VMD-SABO-BiLSTM,IVS-BiLSTM)的容器水平伸缩策...针对Kubernetes默认水平伸缩策略在高并发场景下因间歇时间而导致集群规模无法及时扩展,进而易引发集群性能下降甚至宕机的问题,提出了一种基于加权变分减法双向长短期记忆网络模型(INFO-VMD-SABO-BiLSTM,IVS-BiLSTM)的容器水平伸缩策略。该策略通过将负载预测值输入Pod水平伸缩器(Horizontal Pod Autoscaler,HPA)进行主动扩容,提升了集群对负载变化的感知能力。实验结果表明,所提的IVS-BiLSTM混合预测模型在平均绝对误差(MAE)、均方根误差(RMSE)和平均绝对百分比误差(MAPE)3项指标上分别减少到LSTM的44.29%、44.34%和43.03%,并且较其他主流算法在精度上有显著提升;改进的容器水平伸缩策略在相同负载条件下能够提前扩容,证明了该策略在生产中的可行性。展开更多
文摘准确的高铁沿线风速预测是铁路灾害预警系统的基础需求,为了提升应对和处理强风灾害致突发事件的能力,提出一种基于减法平均优化(subtraction average based optimizer,SABO)算法优化长短时记忆(long short-term memory,LSTM)神经网络的高铁沿线短期风速预测方法。首先,针对风速非线性和非平稳特性,采用极小化极大(min-max,MM)方法对风速数据进行归一化处理;其次,采用SABO算法中的“-v”方法对LSTM模型的关键参数搜索寻优,并构建风速预测模型;最后,以中国宝兰高铁沿线风速采集点采集的实测风速数据为例,对模型进行有效性检验。实验结果表明:SABO算法的寻优效果更加良好,预测精度更高,所建模型的平均绝对误差(mean absolute error,MAE)、平均绝对百分比误差(mean absolute percentage error,MAPE)和均方根误差(route mean square error,RMSE)分别仅为11.96%、1.23%和16.47%,决定系数(r-square,R^(2))为0.995。与其他模型相比,通过SABO算法优化后的LSTM神经网络在短期风速预测上具有较好的拟合效果和更高的预测精度,可为高铁沿线大风预测预警提供一种新的方法和思路。
文摘提出一种基于SABO-GRU-Attention(subtraction average based optimizer-gate recurrent unitattention)的锂电池SOC(state of charge)估计方法。采用基于平均减法优化算法自适应更新GRU神经网络的超参数,融合SE(squeeze and excitation)注意力机制自适应分配各通道权重,提高学习效率。对马里兰大学电池数据集进行预处理,输入电压、电流参数,进行锂电池充放电仿真实验,并搭建锂电池荷电状态实验平台进行储能锂电池充放电实验。结果表明,提出的SOC神经网络估计模型明显优于LSTM、GRU以及PSO-GRU等模型,具有较高的估计精度与应用价值。
基金supported by the National Natural Science Foundation of China (Grant No.51209195)Foundation of Key Laboratory of Mountain Hazards and Earth Surface Process,Chinese Academy of Sciences,Science and Technology Service Network Initiative of Chinese Academy of Sciences (Grant No.KFJ-EW-STS-094)the Youth Foundation of the Institute of Mountain Hazards and Environment,CAS (Grant No.SDS-QN-1302)
文摘Debris flows are one of the common natural hazards in mountainous areas.They often cause devastating damage to the lives and property of local people.The sabo dam construction along a debris flow valley is considered to be a useful method for hazard mitigation.Previous work has concentrated on the different types of sabo dams such as close-type sabo dam,open-type sabo dam.However,little attention has been paid to the spillway structure of sabo dam.In the paper,a new type of spillway structure with lateral contraction was proposed.Debris flow patterns under four different spillway structures were investigated.The projection theory was employed to predict trajectory of debris flow out from the spillway and to estimate the incident angle and terminal velocity before it plunged into the scour hole behind the sabo dam.The results indicated that the estimated data were in good agreement with the experimental ones.The discrepancy between the estimated and experimental values of main parameters remained below 21.82%(relative error).Additionally,the effects of debris flow scales under different spillway structures were considered to study the scour law.Although the debris flow pattern and scour law behind the sabo dam under different operating conditions was analyzed in this paper,further study on the scour mechanism andthe maximum scour depth estimation based on scour theory is still required in the future.
基金the National Natural Science Foundation of China (Nos. 40901007, 50979103)
文摘The erosion shape and the law of development of debris flow sabo dam downstream is a weak part in the study on debris flow erosion. The shape and development of scour pit have an important effect on the stability and safety of debris flow sabo dam, which determines the foundational depth of the dam and the design of protective measures downstream. Study on the scouring law of sabo dam downstream can evaluate the erosion range and reasonably arrange auxiliary protective engineering. Therefore, a series of flume experiments are carried out including different debris flow characteristics （density is varying from 1.5 t/m3 to 2.1 t/m~） and different gully longitudinal slopes. The result shows that the scour pit appears as an oval shape in a plane and deep in the middle while superficial at the ends in the longitudinal section, the position of the maximum depth point moves towards downstream with an increase of flume slope angle. The maximum depth of scour pit is mainly affected by the longitudinal slope of gully, density of debris flow, and the characteristics of gully composition （particle size and the viscosity of soil）. The result also indicates that the viscosity of soil will weaken the erosion extent. The interior slopes of scour pit are different between the upstream and the downstream, and the downstream slope is smaller than the upper one. For the viscous and non-viscous sands with the same distribution of gradation, the interior slope of non- viscous sand is smaller than the viscous sand.According to tbe regression analysis on the experimental data, the quantitative relationship between the interior slope of scour pit, slope of repose under water and the longitudinal slope of gully is established and it can be used to calculate the interior slope of scour pit. The results can provide the basis for the parameter design of the debris flow control engineering foundation.
文摘In order to improve the accuracy of wind turbine fault diagnosis,a wind turbine fault diagnosis method based on Subtraction-Average-Based Optimizer(SABO)optimized Variational Mode Decomposition(VMD)and Kernel Extreme Learning Machine(KELM)is proposed.Firstly,the SABO algorithm was used to optimize the VMD parameters and decompose the original signal to obtain the best modal components,and then the nine features were calculated to obtain the feature vectors.Secondly,the SABO algorithm was used to optimize the KELM parameters,and the training set and the test set were divided according to different proportions.The results were compared with the optimized model without SABO algorithm.The experimental results show that the fault diagnosis method of wind turbine based on SABO-VMD-KELM model can achieve fault diagnosis quickly and effectively,and has higher accuracy.
文摘针对Kubernetes默认水平伸缩策略在高并发场景下因间歇时间而导致集群规模无法及时扩展,进而易引发集群性能下降甚至宕机的问题,提出了一种基于加权变分减法双向长短期记忆网络模型(INFO-VMD-SABO-BiLSTM,IVS-BiLSTM)的容器水平伸缩策略。该策略通过将负载预测值输入Pod水平伸缩器(Horizontal Pod Autoscaler,HPA)进行主动扩容,提升了集群对负载变化的感知能力。实验结果表明,所提的IVS-BiLSTM混合预测模型在平均绝对误差(MAE)、均方根误差(RMSE)和平均绝对百分比误差(MAPE)3项指标上分别减少到LSTM的44.29%、44.34%和43.03%,并且较其他主流算法在精度上有显著提升;改进的容器水平伸缩策略在相同负载条件下能够提前扩容,证明了该策略在生产中的可行性。
