Analysis of the causes of primary revision after unicompartmental knee arthroplasty: A case series

BACKGROUND Unicompartmental knee arthroplasty(UKA)has great advantages in the treatment of unicompartmental knee osteoarthritis,but its revision rate is higher than that of total knee arthroplasty.AIM To summarize and analyse the causes of revision after UKA.METHODS This is a retrospective case series study in which the reasons for the first revision after UKA are summarized.We analysed the clinical symptoms,medical histories,laboratory test results,imaging examination results and treatment processes of the patients who underwent revision and summarized the reasons for primary revision after UKA.RESULTS A total of 13 patients,including 3 males and 10 females,underwent revision surgery after UKA.The average age of the included patients was 67.62 years.The prosthesis was used for 3 d to 72 months.The main reasons for revision after UKA were improper suturing of the surgical opening(1 patient),osteophytes(2 patients),intra-articular loose bodies(2 patients),tibial prosthesis loosening(2 patients),rheumatoid arthritis(1 patient),gasket dislocation(3 patients),anterior cruciate ligament injury(1 patient),and medial collateral ligament injury with residual bone cement(1 patient).CONCLUSION The causes of primary revision after UKA were gasket dislocation,osteophytes,intra-articular loose bodies and tibial prosthesis loosening.Avoidance of these factors may greatly reduce the rate of revision after UKA,improve patient satisfaction and reduce medical burden.

Time series prediction of reservoir bank landslide failure probability considering the spatial variability of soil properties

Historically,landslides have been the primary type of geological disaster worldwide.Generally,the stability of reservoir banks is primarily affected by rainfall and reservoir water level fluctuations.Moreover,the stability of reservoir banks changes with the long-term dynamics of external disastercausing factors.Thus,assessing the time-varying reliability of reservoir landslides remains a challenge.In this paper,a machine learning(ML)based approach is proposed to analyze the long-term reliability of reservoir bank landslides in spatially variable soils through time series prediction.This study systematically investigated the prediction performances of three ML algorithms,i.e.multilayer perceptron(MLP),convolutional neural network(CNN),and long short-term memory(LSTM).Additionally,the effects of the data quantity and data ratio on the predictive power of deep learning models are considered.The results show that all three ML models can accurately depict the changes in the time-varying failure probability of reservoir landslides.The CNN model outperforms both the MLP and LSTM models in predicting the failure probability.Furthermore,selecting the right data ratio can improve the prediction accuracy of the failure probability obtained by ML models.

Defect Detection Model Using Time Series Data Augmentation and Transformation

Time-series data provide important information in many fields,and their processing and analysis have been the focus of much research.However,detecting anomalies is very difficult due to data imbalance,temporal dependence,and noise.Therefore,methodologies for data augmentation and conversion of time series data into images for analysis have been studied.This paper proposes a fault detection model that uses time series data augmentation and transformation to address the problems of data imbalance,temporal dependence,and robustness to noise.The method of data augmentation is set as the addition of noise.It involves adding Gaussian noise,with the noise level set to 0.002,to maximize the generalization performance of the model.In addition,we use the Markov Transition Field(MTF)method to effectively visualize the dynamic transitions of the data while converting the time series data into images.It enables the identification of patterns in time series data and assists in capturing the sequential dependencies of the data.For anomaly detection,the PatchCore model is applied to show excellent performance,and the detected anomaly areas are represented as heat maps.It allows for the detection of anomalies,and by applying an anomaly map to the original image,it is possible to capture the areas where anomalies occur.The performance evaluation shows that both F1-score and Accuracy are high when time series data is converted to images.Additionally,when processed as images rather than as time series data,there was a significant reduction in both the size of the data and the training time.The proposed method can provide an important springboard for research in the field of anomaly detection using time series data.Besides,it helps solve problems such as analyzing complex patterns in data lightweight.

Reservoir characteristics and formation model of Upper Carboniferous bauxite series in eastern Ordos Basin,NW China

Through core observation,thin section identification,X-ray diffraction analysis,scanning electron microscopy,and low-temperature nitrogen adsorption and isothermal adsorption experiments,the lithology and pore characteristics of the Upper Carboniferous bauxite series in eastern Ordos Basin were analyzed to reveal the formation and evolution process of the bauxite reservoirs.A petrological nomenclature and classification scheme for bauxitic rocks based on three units(aluminum hydroxides,iron minerals and clay minerals)is proposed.It is found that bauxitic mudstone is in the form of dense massive and clastic structures,while the(clayey)bauxite is of dense massive,pisolite,oolite,porous soil and clastic structures.Both bauxitic mudstone and bauxite reservoirs develop dissolution pores,intercrystalline pores,and microfractures as the dominant gas storage space,with the porosity less than 10% and mesopores in dominance.The bauxite series in the North China Craton can be divided into five sections,i.e.,ferrilite(Shanxi-style iron ore,section A),bauxitic mudstone(section B),bauxite(section C),bauxite mudstone(debris-containing,section D)and dark mudstone-coal section(section E).The burrow/funnel filling,lenticular,layered/massive bauxite deposits occur separately in the karst platforms,gentle slopes and low-lying areas.The karst platforms and gentle slopes are conducive to surface water leaching,with strong karstification,well-developed pores,large reservoir thickness and good physical properties,but poor strata continuity.The low-lying areas have poor physical properties but relatively continuous and stable reservoirs.The gas enrichment in bauxites is jointly controlled by source rock,reservoir rock and fractures.This recognition provides geological basis for the exploration and development of natural gas in the Upper Carboniferous in the study area and similar bauxite systems.

Cross-Dimension Attentive Feature Fusion Network for Unsupervised Time-Series Anomaly Detection

Time series anomaly detection is crucial in various industrial applications to identify unusual behaviors within the time series data.Due to the challenges associated with annotating anomaly events,time series reconstruction has become a prevalent approach for unsupervised anomaly detection.However,effectively learning representations and achieving accurate detection results remain challenging due to the intricate temporal patterns and dependencies in real-world time series.In this paper,we propose a cross-dimension attentive feature fusion network for time series anomaly detection,referred to as CAFFN.Specifically,a series and feature mixing block is introduced to learn representations in 1D space.Additionally,a fast Fourier transform is employed to convert the time series into 2D space,providing the capability for 2D feature extraction.Finally,a cross-dimension attentive feature fusion mechanism is designed that adaptively integrates features across different dimensions for anomaly detection.Experimental results on real-world time series datasets demonstrate that CAFFN performs better than other competing methods in time series anomaly detection.

Data-augmented landslide displacement prediction using generative adversarial network

Landslides are destructive natural disasters that cause catastrophic damage and loss of life worldwide.Accurately predicting landslide displacement enables effective early warning and risk management.However,the limited availability of on-site measurement data has been a substantial obstacle in developing data-driven models,such as state-of-the-art machine learning(ML)models.To address these challenges,this study proposes a data augmentation framework that uses generative adversarial networks(GANs),a recent advance in generative artificial intelligence(AI),to improve the accuracy of landslide displacement prediction.The framework provides effective data augmentation to enhance limited datasets.A recurrent GAN model,RGAN-LS,is proposed,specifically designed to generate realistic synthetic multivariate time series that mimics the characteristics of real landslide on-site measurement data.A customized moment-matching loss is incorporated in addition to the adversarial loss in GAN during the training of RGAN-LS to capture the temporal dynamics and correlations in real time series data.Then,the synthetic data generated by RGAN-LS is used to enhance the training of long short-term memory(LSTM)networks and particle swarm optimization-support vector machine(PSO-SVM)models for landslide displacement prediction tasks.Results on two landslides in the Three Gorges Reservoir(TGR)region show a significant improvement in LSTM model prediction performance when trained on augmented data.For instance,in the case of the Baishuihe landslide,the average root mean square error(RMSE)increases by 16.11%,and the mean absolute error(MAE)by 17.59%.More importantly,the model’s responsiveness during mutational stages is enhanced for early warning purposes.However,the results have shown that the static PSO-SVM model only sees marginal gains compared to recurrent models such as LSTM.Further analysis indicates that an optimal synthetic-to-real data ratio(50%on the illustration cases)maximizes the improvements.This also demonstrates the robustness and effectiveness of supplementing training data for dynamic models to obtain better results.By using the powerful generative AI approach,RGAN-LS can generate high-fidelity synthetic landslide data.This is critical for improving the performance of advanced ML models in predicting landslide displacement,particularly when there are limited training data.Additionally,this approach has the potential to expand the use of generative AI in geohazard risk management and other research areas.

Improved Responses with Multitaper Spectral Analysis for Magnetotelluric Time Series Data Processing:Examples from Field Data

In order to attain good quality transfer function estimates from magnetotelluric field data(i.e.,smooth behavior and small uncertainties across all frequencies),we compare time series data processing with and without a multitaper approach for spectral estimation.There are several common ways to increase the reliability of the Fourier spectral estimation from experimental(noisy)data;for example to subdivide the experimental time series into segments,taper these segments(using single taper),perform the Fourier transform of the individual segments,and average the resulting spectra.

Unsupervised Time Series Segmentation: A Survey on Recent Advances

Time series segmentation has attracted more interests in recent years,which aims to segment time series into different segments,each reflects a state of the monitored objects.Although there have been many surveys on time series segmentation,most of them focus more on change point detection(CPD)methods and overlook the advances in boundary detection(BD)and state detection(SD)methods.In this paper,we categorize time series segmentation methods into CPD,BD,and SD methods,with a specific focus on recent advances in BD and SD methods.Within the scope of BD and SD,we subdivide the methods based on their underlying models/techniques and focus on the milestones that have shaped the development trajectory of each category.As a conclusion,we found that:(1)Existing methods failed to provide sufficient support for online working,with only a few methods supporting online deployment;(2)Most existing methods require the specification of parameters,which hinders their ability to work adaptively;(3)Existing SD methods do not attach importance to accurate detection of boundary points in evaluation,which may lead to limitations in boundary point detection.We highlight the ability to working online and adaptively as important attributes of segmentation methods,the boundary detection accuracy as a neglected metrics for SD methods.

ESTIMATION OF AVERAGE DIFFERENTIAL ENTROPY FOR A STATIONARY ERGODIC SPACE-TIME RANDOM FIELD ON A BOUNDED AREA

In this paper,we mainly discuss a discrete estimation of the average differential entropy for a continuous time-stationary ergodic space-time random field.By estimating the probability value of a time-stationary random field in a small range,we give an entropy estimation and obtain the average entropy estimation formula in a certain bounded space region.It can be proven that the estimation of the average differential entropy converges to the theoretical value with a probability of 1.In addition,we also conducted numerical experiments for different parameters to verify the convergence result obtained in the theoretical proofs.

Determining the planting year of navel orange trees in mountainous and hilly areas of southern China:a remote sensing based method

Remote sensing has demonstrated validity in determining the planting year of deciduous fruit trees;however,its effectiveness in ascertaining the planting year of evergreen fruit trees remains unverified.Furthermore,the sources of error associated with using remote sensing to determine the planting year of fruit trees remain unclear.This study investigates several cultivated sweet orange(Citrus sinensis)varieties,which are extensively cultivated throughout subtropical China.We analyzed Landsat time series data from 132 navel orange orchards in Gannan,covering the period from 1993 to 2021.For each orchard,Google Earth Engine was employed to extract three vegetation indices—Enhanced Vegetation Index(EVI),Normalized Difference Vegetation Index(NDVI),and Normalized Burn Ratio(NBR)—for each available date,thereby generating three distinct vegetation index time series.The planting year of navel orange trees was identified based on abrupt changes observed in these time series.The principal sources of error in determining the planting year were investigated using the Wilcoxon signed-rank test,Spearman's correlation analysis,and Kruskal-Wallis H test.Key findings include:(1)Following the planting of navel orange trees,EVI,NDVI,and NBR exhibited fluctuations and a gradual increase over time,peaking approximately 10 to 15 years later.(2)The vegetation index time series derived from Landsat imagery effectively determined the planting year of evergreen navel orange trees in orchards,even within highly fragmented landscapes.Among these indices,NDVI and NBR time series outperformed the EVI time series.Specifically,the average determination errors for EVI,NDVI,and NBR time series were 6.4,1.8,and 2.8 years,respectively.(3)Major sources of error included the methods used to construct the time series,the selection of vegetation indices,and the orchard management practices.Overall,this study provides a viable method for determining the planting year of evergreen navel orange trees in fragmented landscapes and offers insights into factors contributing to uncertainty in planting year determination.

Reconstructing historical forest fire risk in the non-satellite era using the improved forest fire danger index and long short-term memory deep learning-a case study in Sichuan Province,southwestern China

Historical forest fire risk databases are vital for evaluating the effectiveness of past forest management approaches,enhancing forest fire warnings and emergency response capabilities,and accurately budgeting potential carbon emissions resulting from fires.However,due to the unavailability of spatial information technology,such databases are extremely difficult to build reliably and completely in the non-satellite era.This study presented an improved forest fire risk reconstruction framework that integrates a deep learning-based time series prediction model and spatial interpolation to address the challenge in Sichuan Province,southwestern China.First,the forest fire danger index(FFDI)was improved by supplementing slope and aspect information.We compared the performances of three time series models,namely,the autoregressive integrated moving average(ARIMA),Prophet and long short-term memory(LSTM)in predicting the modified forest fire danger index(MFFDI).The bestperforming model was used to retrace the MFFDI of individual stations from 1941 to 1970.Following this,the Anusplin spatial interpolation method was used to map the distributions of the MFFDI at five-year intervals,which were then subjected to weighted overlay with the distance-to-river layer to generate forest fire risk maps for reconstructing the forest fire danger database.The results revealed LSTM as the most accurate in fitting and predicting the historical MFFDI,with a fitting determination coefficient(R^2)of 0.709,mean square error(MSE)of0.047,and validation R^2 and MSE of 0.508 and 0.11,respectively.Independent validation of the predicted forest fire risk maps indicated that 5 out of 7 historical forest fire events were located in forest fire-prone areas,which is higher than the results determined from the original FFDI(2 out of 7).This proves the effectiveness of the improved MFFDI and indicates a high level of reliability of the historical forest fire risk reconstruction method proposed in this study.

Periodic signal extraction of GNSS height time series based on adaptive singular spectrum analysis

Singular spectrum analysis is widely used in geodetic time series analysis.However,when extracting time-varying periodic signals from a large number of Global Navigation Satellite System(GNSS)time series,the selection of appropriate embedding window size and principal components makes this method cumbersome and inefficient.To improve the efficiency and accuracy of singular spectrum analysis,this paper proposes an adaptive singular spectrum analysis method by combining spectrum analysis with a new trace matrix.The running time and correlation analysis indicate that the proposed method can adaptively set the embedding window size to extract the time-varying periodic signals from GNSS time series,and the extraction efficiency of a single time series is six times that of singular spectrum analysis.The method is also accurate and more suitable for time-varying periodic signal analysis of global GNSS sites.

Performance Degradation Prediction of Proton Exchange Membrane Fuel Cell Based on CEEMDAN-KPCA and DA-GRU Networks

In order to improve the performance degradation prediction accuracy of proton exchange membrane fuel cell(PEMFC),a fusion prediction method(CKDG)based on adaptive noise complete ensemble empirical mode decomposition(CEEMDAN),kernel principal component analysis(KPCA)and dual attention mechanism gated recurrent unit neural network(DA-GRU)was proposed.CEEMDAN and KPCA were used to extract the input feature data sequence,reduce the influence of random factors,and capture essential feature components to reduce the model complexity.The DA-GRU network helps to learn the feature mapping relationship of data in long time series and predict the changing trend of performance degradation data more accurately.The actual aging experimental data verify the performance of the CKDG method.The results show that under the steady-state condition of 20%training data prediction,the CKDA method can reduce the root mean square error(RMSE)by 52.7%and 34.6%,respectively,compared with the traditional LSTM and GRU neural networks.Compared with the simple DA-GRU network,RMSE is reduced by 15%,and the degree of over-fitting is reduced,which has higher accuracy.It also shows excellent prediction performance under the dynamic condition data set and has good universality.

Phenology of different types of vegetation and their response to climate change in the Qilian Mountains,China

The Qilian Mountains(QM)possess a delicate vegetation ecosystem,amplifying the evident response of vegetation phenology to climate change.The relationship between changes in vegetation growth and climate remains complex.To this end,we used MODIS NDVI data to extract the phenological parameters of the vegetation including meadow(MDW),grassland(GSD),and alpine vegetation(ALV))in the QM from 2002 to 2021.Then,we employed path analysis to reveal the direct and indirect impacts of seasonal climate change on vegetation phenology.Additionally,we decomposed the vegetation phenology in a time series using the trigonometric seasonality,Box-Cox transformation,ARMA errors,and Trend Seasonal components model(TBATS).The findings showed a distinct pattern in the vegetation phenology of the QM,characterized by a progressive shift towards an earlier start of the growing season(SOS),a delayed end of the growing season(EOS),and an extended length of the growing season(LOS).The growth cycle of MDW,GSD,and ALV in the QM species is clearly defined.The SOS for MDW and GSD occurred earlier,mainly between late April and August,while the SOS for ALVs occurred between mid-May and mid-August,a one-month delay compared to the other vegetation.The EOS in MDW and GSD were concentrated between late August and April and early September and early January,respectively.Vegetation phenology exhibits distinct responses to seasonal temperature and precipitation patterns.The advancement and delay of SOS were mainly influenced by the direct effect of spring temperatures and precipitation,which affected 19.59%and 22.17%of the study area,respectively.The advancement and delay of EOS were mainly influenced by the direct effect of fall temperatures and precipitation,which affected 30.18%and 21.17%of the area,respectively.On the contrary,the direct effects of temperature and precipitation in summer and winter on vegetation phenology seem less noticeable and were mainly influenced by indirect effects.The indirect effect of winter precipitation is the main factor affecting the advance or delay of SOS,and the area proportions were 16.29%and 23.42%,respectively.The indirect effects of fall temperatures and precipitation were the main factors affecting the delay and advancement of EOS,respectively,with an area share of 15.80%and 21.60%.This study provides valuable insight into the relationship between vegetation phenology and climate change,which can be of great practical value for the ecological protection of the Qinghai-Tibetan Plateau as well as for the development of GSD ecological animal husbandry in the QM alpine pastoral area.

Probability Distribution Characteristics of Strong Nonlinear Waves Under Typhoon Conditions in the Northern South China Sea

The generation and propagation mechanism of strong nonlinear waves in the South China Sea is an essential research area. In this study, the third-generation wave model WAVEWATCH Ⅲ is employed to simulate wave fields under extreme sea states. The model, integrating the ST6 source term, is validated against observed data, demonstrating its credibility. The spatial distribution of the occurrence probability of strong nonlinear waves during typhoons is shown, and the waves in the straits and the northeastern part of the South China Sea show strong nonlinear characteristics. The high-order spectral model HOS-ocean is employed to simulate the random wave surface series beneath five different platform areas. The waves during the typhoon exhibit strong nonlinear characteristics, and freak waves exist. The space-varying probability model is established to describe the short-term probability distribution of nonlinear wave series. The exceedance probability distributions of the wave surface beneath different platform areas are compared and analyzed. The results show that with an increase in the platform area, the probability of a strong nonlinear wave beneath the platform increases.

Nitrogen isotope stratigraphy of the Early Cambrian successions in the Tarim Basin:Spatial variability of nitrogen cycling and its implication for paleo-oceanic redox conditions

The Early Cambrian represents a critical time period characterized by extraordinary biological innovations and dynamic redox conditions in seawaters.Nitrogen isotopic signatures of ancient sediments have the potential to elucidate the evolutionary path of marine redox states and the biogeochemical nitrogen cycle within the water column of the Early Cambrian ocean.While existing research on this topic has predominantly focused on South China,the exploration of other continental margins has been limited,leaving contradictory hypotheses untested.In this study,pairedδ^(15)N andδ^(13)C org analyses were performed on the Lower Cambrian successions from the Shiairike section(inner ramp)and Well Tadong 2(deep shelf/basin)in the northwestern and eastern Tarim Basin,respectively.Our data from the Shiairike section reveal a discernible shift in the operation of different nitrogen cycles for the black chert-shale unit,also referred to as the black rock series in Chinese literature,of the Yurtus Formation(Fortunian stage to lower Stage 3).Oscillatingδ^(15)N values for its lower part are suggestive of alternating anaerobic assimilation of NH 4+and denitrification/anammox.This is likely attributed to a shallow,unstable chemocline consistent with the upwelling and incursion of deep,anoxic waters during a major transgression.In contrast,aerobic nitrogen cycling,indicated by positiveδ^(15)N values of>2‰,dominated the upper part alongside a reduction in upwelling intensity.On the other hand,theδ^(15)N signatures of Xishanbulake and Xidashan Formations of Well Tadong 2,which encompass a time interval from the Cambrian Fortunian Age to Age 4,are indicative of N_(2)fixation by diazotrophs as the major nitrogen source.The two studied intervals,although not time-equivalent,exhibit separated states of nitrogen cycling at least during the deposition of the Yurtus black rock series.The spatially different nitrogen cycling of the studied sections is compatible with a redox-stratified ocean during the deposition of the Yurtus black rock series.The build-up of a NO_(3)−reservoir and aerobic nitrogen cycling in seawater was largely restricted to near-shore settings whereas anaerobic nitrogen cycling dominated by N_(2)fixation served as the main nitrogen uptake pathway in off-shore settings.

Three-Dimensional Multiferroic Structures under Time-Harmonic Loading

Magneto-electro-elastic(MEE)materials are a specific class of advanced smart materials that simultaneouslymanifest the coupling behavior under electric,magnetic,and mechanical loads.This unique combination ofproperties allows MEE materials to respond to mechanical,electric,and magnetic stimuli,making them versatile forvarious applications.This paper investigates the static and time-harmonic field solutions induced by the surface loadin a three-dimensional(3D)multilayered transversally isotropic(TI)linear MEE layered solid.Green’s functionscorresponding to the applied uniform load(in both horizontal and vertical directions)are derived using the FourierBessel series(FBS)system of vector functions.By virtue of this FBS method,two sets of first-order ordinarydifferential equations(i.e.,N-type and LM-type)are obtained,with the expansion coefficients being Love numbers.It is noted that the LM-type system corresponds to the MEE-coupled P-,SV-,and Rayleigh waves,while the N-typecorresponds to the purely elastic SH-and Love waves.By applying the continuity conditions across interfaces,the solutions for each layer of the structure(from the bottom to the top)are derived using the dual-variable andposition(DVP)method.This method(i.e.,DVP)is unconditionally stable when propagating solutions throughdifferent layers.Numerical examples illustrate the impact of load types,layering,and frequency on the response ofthe structure,as well as the accuracy and convergence of the proposed approach.The numerical results are usefulin designing smart devices made of MEE solids,which are applicable to engineering fields like renewable energy.

An Innovative Deep Architecture for Flight Safety Risk Assessment Based on Time Series Data

With the development of the integration of aviation safety and artificial intelligence,research on the combination of risk assessment and artificial intelligence is particularly important in the field of risk management,but searching for an efficient and accurate risk assessment algorithm has become a challenge for the civil aviation industry.Therefore,an improved risk assessment algorithm(PS-AE-LSTM)based on long short-term memory network(LSTM)with autoencoder(AE)is proposed for the various supervised deep learning algorithms in flight safety that cannot adequately address the problem of the quality on risk level labels.Firstly,based on the normal distribution characteristics of flight data,a probability severity(PS)model is established to enhance the quality of risk assessment labels.Secondly,autoencoder is introduced to reconstruct the flight parameter data to improve the data quality.Finally,utilizing the time-series nature of flight data,a long and short-termmemory network is used to classify the risk level and improve the accuracy of risk assessment.Thus,a risk assessment experimentwas conducted to analyze a fleet landing phase dataset using the PS-AE-LSTMalgorithm to assess the risk level associated with aircraft hard landing events.The results show that the proposed algorithm achieves an accuracy of 86.45%compared with seven baseline models and has excellent risk assessment capability.

Deformation monitoring of long-span railway bridges based on SBAS-InSAR technology

The deformation monitoring of long-span railway bridges is significant to ensure the safety of human life and property.The interferometric synthetic aperture radar(In SAR)technology has the advantage of high accuracy in bridge deformation monitoring.This study monitored the deformation of the Ganjiang Super Bridge based on the small baseline subsets(SBAS)In SAR technology and Sentinel-1A data.We analyzed the deformation results combined with bridge structure,temperature,and riverbed sediment scouring.The results are as follows:(1)The Ganjiang Super Bridge area is stable overall,with deformation rates ranging from-15.6 mm/yr to 10.7 mm/yr(2)The settlement of the Ganjiang Super Bridge deck gradually increases from the bridge tower toward the main span,which conforms to the typical deformation pattern of a cable-stayed bridge.(3)The sediment scouring from the riverbed cause the serious settlement on the bridge’s east side compared with that on the west side.(4)The bridge deformation negatively correlates with temperature,with a faster settlement at a higher temperature and a slow rebound trend at a lower temperature.The study findings can provide scientific data support for the health monitoring of long-span railway bridges.

TSCND:Temporal Subsequence-Based Convolutional Network with Difference for Time Series Forecasting

Time series forecasting plays an important role in various fields, such as energy, finance, transport, and weather. Temporal convolutional networks (TCNs) based on dilated causal convolution have been widely used in time series forecasting. However, two problems weaken the performance of TCNs. One is that in dilated casual convolution, causal convolution leads to the receptive fields of outputs being concentrated in the earlier part of the input sequence, whereas the recent input information will be severely lost. The other is that the distribution shift problem in time series has not been adequately solved. To address the first problem, we propose a subsequence-based dilated convolution method (SDC). By using multiple convolutional filters to convolve elements of neighboring subsequences, the method extracts temporal features from a growing receptive field via a growing subsequence rather than a single element. Ultimately, the receptive field of each output element can cover the whole input sequence. To address the second problem, we propose a difference and compensation method (DCM). The method reduces the discrepancies between and within the input sequences by difference operations and then compensates the outputs for the information lost due to difference operations. Based on SDC and DCM, we further construct a temporal subsequence-based convolutional network with difference (TSCND) for time series forecasting. The experimental results show that TSCND can reduce prediction mean squared error by 7.3% and save runtime, compared with state-of-the-art models and vanilla TCN.