Oceanic turbulence measurements made by an acoustic Doppler velocimeter(ADV)suffer from noise that potentially affects the estimates of turbulence statistics.This study examines the abilities of Kalman filtering and a...Oceanic turbulence measurements made by an acoustic Doppler velocimeter(ADV)suffer from noise that potentially affects the estimates of turbulence statistics.This study examines the abilities of Kalman filtering and autoregressive moving average models to eliminate noise in ADV velocity datasets of laboratory experiments and offshore observations.Results show that the two methods have similar performance in ADV de-noising,and both effectively reduce noise in ADV velocities,even in cases of high noise.They eliminate the noise floor at high frequencies of the velocity spectra,leading to a longer range that effectively fits the Kolmogorov-5/3 slope at midrange frequencies.After de-noising adopting the two methods,the values of the mean velocity are almost unchanged,while the root-mean-square horizontal velocities and thus turbulent kinetic energy decrease appreciably in these experiments.The Reynolds stress is also affected by high noise levels,and de-noising thus reduces uncertainties in estimating the Reynolds stress.展开更多
An absolute gravimeter is a precision instrument for measuring gravitational acceleration, which plays an important role in earthquake monitoring, crustal deformation, national defense construction, etc. The frequency...An absolute gravimeter is a precision instrument for measuring gravitational acceleration, which plays an important role in earthquake monitoring, crustal deformation, national defense construction, etc. The frequency of laser interference fringes of an absolute gravimeter gradually increases with the fall time. Data are sparse in the early stage and dense in the late stage. The fitting accuracy of gravitational acceleration will be affected by least-squares fitting according to the fixed number of zero-crossing groups. In response to this problem, a method based on Fourier series fitting is proposed in this paper to calculate the zero-crossing point. The whole falling process is divided into five frequency bands using the Hilbert transformation. The multiplicative auto-regressive moving average model is then trained according to the number of optimal zero-crossing groups obtained by the honey badger algorithm. Through this model, the number of optimal zero-crossing groups determined in each segment is predicted by the least-squares fitting. The mean value of gravitational acceleration in each segment is then obtained. The method can improve the accuracy of gravitational measurement by more than 25% compared to the fixed zero-crossing groups method. It provides a new way to improve the measuring accuracy of an absolute gravimeter.展开更多
In the paper,the autoregressive moving average model for matrix time series(MARMA)is inves-tigated.The properties of the MARMA model are investigated by using the conditional least square estimation,the conditional ma...In the paper,the autoregressive moving average model for matrix time series(MARMA)is inves-tigated.The properties of the MARMA model are investigated by using the conditional least square estimation,the conditional maximum likelihood estimation,the projection theorem in Hilbert space and the decomposition technique of time series,which include necessary and suf-ficient conditions for stationarity and invertibility,model parameter estimation,model testing and model forecasting.展开更多
The stock market is a vital component of the broader financial system,with its dynamics closely linked to economic growth.The challenges associated with analyzing and forecasting stock prices have persisted since the ...The stock market is a vital component of the broader financial system,with its dynamics closely linked to economic growth.The challenges associated with analyzing and forecasting stock prices have persisted since the inception of financial markets.By examining historical transaction data,latent opportunities for profit can be uncovered,providing valuable insights for both institutional and individual investors to make more informed decisions.This study focuses on analyzing historical transaction data from four banks to predict closing price trends.Various models,including decision trees,random forests,and Long Short-Term Memory(LSTM)networks,are employed to forecast stock price movements.Historical stock transaction data serves as the input for training these models,which are then used to predict upward or downward stock price trends.The study’s empirical results indicate that these methods are effective to a degree in predicting stock price movements.The LSTM-based deep neural network model,in particular,demonstrates a commendable level of predictive accuracy.This conclusion is reached following a thorough evaluation of model performance,highlighting the potential of LSTM models in stock market forecasting.The findings offer significant implications for advancing financial forecasting approaches,thereby improving the decision-making capabilities of investors and financial institutions.展开更多
计算流体动力学与刚体动力学(Computational Fluid Dynamics and Rigid Body Dynamics,CFD/RBD)耦合仿真是旋转弹飞行性能评估的常用方法之一,但由于需要进行大量CFD计算,该方法效率较低。建立一个高效、精确且泛化能力强的气动力模型...计算流体动力学与刚体动力学(Computational Fluid Dynamics and Rigid Body Dynamics,CFD/RBD)耦合仿真是旋转弹飞行性能评估的常用方法之一,但由于需要进行大量CFD计算,该方法效率较低。建立一个高效、精确且泛化能力强的气动力模型并以之替代耦合仿真中的CFD模块,可以大幅度提升仿真效率。针对前述旋转弹气动力建模问题,提出一种结合系统辨识和迁移学习的建模方法。给定旋转弹运动初始条件并采用CFD/RBD耦合仿真获得样本,采用自回归滑动平均方法建立原始气动力模型,同时采用长短时记忆网络建立状态预测模型。当初始条件变化不大时,原始气动力模型仍然适用;当初始条件发生较大改变时,利用迁移学习将状态预测模型迁移到该初始条件下,并预测相应初始条件下的状态参数,基于预测得到的状态参数,采用自回归滑动平均方法建立气动力模型。算例结果表明:所提方法适用于初始转速和俯仰角变化较大时对旋转弹气动力的精确建模;与直接以CFD/RBD耦合仿真结果为样本、采用自回归滑动平均方法建模相比,在精度相同时建模时间缩短了一半。展开更多
交通流量因受周期性特征、突发状况等多重因素影响,现有模型的预测精度无法满足实际要求.对此,本文提出了基于误差补偿的多模态协同交通流预测模型(Multimodal Collaborative traffic flow prediction model based on Error Compensatio...交通流量因受周期性特征、突发状况等多重因素影响,现有模型的预测精度无法满足实际要求.对此,本文提出了基于误差补偿的多模态协同交通流预测模型(Multimodal Collaborative traffic flow prediction model based on Error Compensation,MCEC).针对传统预测模型不能兼顾时间序列和协变量的问题,提出基于小波分析的特征拓展方法,该方法引入聚类算法得到节假日标签特征,将拥堵指数、交通事故图、天气信息作为拓展特征,对特征进行多尺度分解.在训练阶段,为达到充分学习各部分数据、最优匹配模型的效果,采用差分整合移动平均自回归模型(Autoreg Ressive Integrated Moving Average Model,ARIMA)、长短期记忆神经网络(Long Short-Term Memory network,LSTM)、限制动态时间规整技术(Dynamic Time Warping,DTW)以及自注意力机制(Self-Attention),设计了多模态协同模型训练.在误差补偿阶段,将得到的相应过程值输入基于支持向量机回归(Support Vector Regression,SVR)的误差补偿模块,对各分量的误差进行学习、补偿,并重构得到预测结果.使用公开的高速公路数据集对MCEC进行验证,在多个时间间隔下对比实验结果表明,MCEC在交通流量预测中的平均绝对百分比误差(Mean Absolute Percentage Error,MAPE)达到17.02%,比LSTM-SVR、ConvLSTM(Convolutional Long Short-Term Memory network)、ST-GCN(Spatial Temporal Graph Convolutional Networks)、MFFB(Multi-stream Feature Fusion Block)、Transformer等预测模型具有更高的预测精度,MCEC模型具有较好的有效性与合理性.展开更多
基金The National Key Research and Development Program of China under contract No.2017YFC1404000the Basic Scientific Fund for National Public Research Institutes of China under contract No.2018S03the National Natural Science Foundation of China under contract Nos 41776038 and 41821004
文摘Oceanic turbulence measurements made by an acoustic Doppler velocimeter(ADV)suffer from noise that potentially affects the estimates of turbulence statistics.This study examines the abilities of Kalman filtering and autoregressive moving average models to eliminate noise in ADV velocity datasets of laboratory experiments and offshore observations.Results show that the two methods have similar performance in ADV de-noising,and both effectively reduce noise in ADV velocities,even in cases of high noise.They eliminate the noise floor at high frequencies of the velocity spectra,leading to a longer range that effectively fits the Kolmogorov-5/3 slope at midrange frequencies.After de-noising adopting the two methods,the values of the mean velocity are almost unchanged,while the root-mean-square horizontal velocities and thus turbulent kinetic energy decrease appreciably in these experiments.The Reynolds stress is also affected by high noise levels,and de-noising thus reduces uncertainties in estimating the Reynolds stress.
基金Project supported by the National Key R&D Program of China (Grant No. 2022YFF0607504)。
文摘An absolute gravimeter is a precision instrument for measuring gravitational acceleration, which plays an important role in earthquake monitoring, crustal deformation, national defense construction, etc. The frequency of laser interference fringes of an absolute gravimeter gradually increases with the fall time. Data are sparse in the early stage and dense in the late stage. The fitting accuracy of gravitational acceleration will be affected by least-squares fitting according to the fixed number of zero-crossing groups. In response to this problem, a method based on Fourier series fitting is proposed in this paper to calculate the zero-crossing point. The whole falling process is divided into five frequency bands using the Hilbert transformation. The multiplicative auto-regressive moving average model is then trained according to the number of optimal zero-crossing groups obtained by the honey badger algorithm. Through this model, the number of optimal zero-crossing groups determined in each segment is predicted by the least-squares fitting. The mean value of gravitational acceleration in each segment is then obtained. The method can improve the accuracy of gravitational measurement by more than 25% compared to the fixed zero-crossing groups method. It provides a new way to improve the measuring accuracy of an absolute gravimeter.
基金This paper is partially supported by the basic scientific research business expenses of Universities in Xinjiang,China[Grant Number XQZX20230057]the National Natural Science Foundation of China[Grant Number 11671142].
文摘In the paper,the autoregressive moving average model for matrix time series(MARMA)is inves-tigated.The properties of the MARMA model are investigated by using the conditional least square estimation,the conditional maximum likelihood estimation,the projection theorem in Hilbert space and the decomposition technique of time series,which include necessary and suf-ficient conditions for stationarity and invertibility,model parameter estimation,model testing and model forecasting.
文摘The stock market is a vital component of the broader financial system,with its dynamics closely linked to economic growth.The challenges associated with analyzing and forecasting stock prices have persisted since the inception of financial markets.By examining historical transaction data,latent opportunities for profit can be uncovered,providing valuable insights for both institutional and individual investors to make more informed decisions.This study focuses on analyzing historical transaction data from four banks to predict closing price trends.Various models,including decision trees,random forests,and Long Short-Term Memory(LSTM)networks,are employed to forecast stock price movements.Historical stock transaction data serves as the input for training these models,which are then used to predict upward or downward stock price trends.The study’s empirical results indicate that these methods are effective to a degree in predicting stock price movements.The LSTM-based deep neural network model,in particular,demonstrates a commendable level of predictive accuracy.This conclusion is reached following a thorough evaluation of model performance,highlighting the potential of LSTM models in stock market forecasting.The findings offer significant implications for advancing financial forecasting approaches,thereby improving the decision-making capabilities of investors and financial institutions.
文摘目的运用自回归积分滑动平均模型(Autoregressive Intergrated Moving Average,ARIMA)建立月平均住院费用和住院日的医学经济学模型,为医院精细化管理提供依据。方法利用R4.0.2软件对2017年1月—2021年12月四川大学华西医院宜宾医院(宜宾市第二人民医院)的平均住院费用和住院日数据建立时间序列ARIMA预测模型。结果住院费用最优模型为ARIMA(0,1,1),赤池信息准则(Akaike information criterion,AIC)=924.35,贝叶斯信息准则(Bayesian Information Criterion,BIC)=928.51,残差Ljung-Box Q=12.51(P=0.768),可认为残差序列为白噪声。平均住院日的最优模型为ARIMA(5,1,1),AIC=87.49,BIC=104.11,残差Ljung-Box Q=10.05(P=0.612),可认为残差序列为白噪声。2022年1—12月实际值与预测值基本吻合,月人均住院费用和人均住院日的平均相对误差为0.55%、0.29%。结论建立基于时间序列ARIMA模型能够为合理配置卫生资源提供强有力的数据支撑。
文摘计算流体动力学与刚体动力学(Computational Fluid Dynamics and Rigid Body Dynamics,CFD/RBD)耦合仿真是旋转弹飞行性能评估的常用方法之一,但由于需要进行大量CFD计算,该方法效率较低。建立一个高效、精确且泛化能力强的气动力模型并以之替代耦合仿真中的CFD模块,可以大幅度提升仿真效率。针对前述旋转弹气动力建模问题,提出一种结合系统辨识和迁移学习的建模方法。给定旋转弹运动初始条件并采用CFD/RBD耦合仿真获得样本,采用自回归滑动平均方法建立原始气动力模型,同时采用长短时记忆网络建立状态预测模型。当初始条件变化不大时,原始气动力模型仍然适用;当初始条件发生较大改变时,利用迁移学习将状态预测模型迁移到该初始条件下,并预测相应初始条件下的状态参数,基于预测得到的状态参数,采用自回归滑动平均方法建立气动力模型。算例结果表明:所提方法适用于初始转速和俯仰角变化较大时对旋转弹气动力的精确建模;与直接以CFD/RBD耦合仿真结果为样本、采用自回归滑动平均方法建模相比,在精度相同时建模时间缩短了一半。
