Actual Values Research Articles

Single vs dual source surface energy balance model based actual evapotranspiration estimation

The current study aims to inter-compare the performance efficiency of the single and the dual source surface energy balance modeling approaches, namely EEFlux and SETMI, respectively for real time catchment scale - crop water demand estimations. For this, the afore-stated two surface energy balance modelling approaches were applied on the Narmada Canal Project, Sanchore, Rajasthan, India for estimating catchment scale actual evapotranspiration (ETa) values for the Rabi cropping seasons of the years 2013-14 and 2018-19, after incorporating the basic satellite data derived inputs viz. Land use, Land surface temperature and Gridded weather data. Due to the non-availability of the catchment scale ground based daily reference evapotranspiration (ETo) values for the study area, the Global Land Data Assimilation System based gridded meteorological data product was utilized, as a substitute for obtaining observed actual evapotranspiration (ETa) values for the investigated Rabi seasons of the study area. These actual evapotranspiration values were compared with those estimated through the single source, EEFlux and the dual source, SETMI modelling approaches to ascertain their comparative performance efficiency through the use of the five statistical indices viz. Mean Absolute Error, Root Mean Square Error, Mean Bias Error, Nash-Sutcliffe Efficiency and the Index of Agreement. The investigations revealed almost at par performance of the two modelling approaches. However, it was concluded that in contrast to the more detailed dual source approach i.e., SETMI, the simple single source approach i.e., EEFlux seemed to be more promising due to its user-friendly implementation and input data automation.

Comparison of force profiles from two musculoskeletal palpation methods: A methodological study.

The Diagnostic Criteria for Temporomandibular Disorders (DC/TMD) protocol recommends a 5 s and 1 kg force dynamic palpation around the lateral condylar pole of the temporomandibular joint. However, the accuracy and precision of the generated force are not known. To assess and compare the force profiles generated from dynamic palpation manually and using a palpometer, based on the forces and time recommendations suggested by the DC/TMD protocol. Nineteen healthy adults applied forces of 0.5 kg, 1.0 kg and 2.0 kg on a calibrated force sensor in a circular motion within target times of 2 s and 5 s. Participants used their right index finger for manual palpation and a calibrated palpometer for device-assisted palpation. Ten repetitions of each target force at both target times were applied. Time taken to complete each application was recorded. Repeated measures analysis of variance was used for analysis of accuracy measured as the relative difference between targeted force and actual force values and precision measured as the coefficient of variation (CV) within the 10 repeated measurements. Accuracy was significantly lower (better) and precision higher (lower CV) with the palpometer than with manual palpation (p < .001). There were significant differences in accuracy and precision between the different forces but not palpation times. Most participants could not achieve the target times and tended to be faster, irrespective of the palpation method (p > .063). A palpometer is a more accurate and precise palpation method for dynamic force assessment compared to manual palpation; however, it remains difficult to standardize the palpation duration.

Solar Energy Prediction Based on Intelligent Predictive Controller Algorithm

The technological advancement in all countries leads to massive energy demand. The energy trading companies struggle daily to meet their customers’ power demands. For a good quality, disturbance-free, and reliable power supply, one must balance electricity generation and consumption at the grid level. There is a profound change in distribution networks due to the intervention of renewable energy generation and grid interactions. Renewable energy sources like solar and wind depend on environmental factors and are subject to unpredictable variations. Earlier, energy distribution companies faced a significant challenge in demand forecasting since it is often unpredictable. With the prediction of the ever-varying power from renewable sources, the power generation and distribution agencies are facing a challenge in supply-side predictions. Several forecasting techniques have evolved, and machine learning techniques like the model predictive controller are suitable for arduous tasks like predicting weather-dependent power generation in advance. This paper employs a Model Predictive Controller (MPC) to predict the solar array’s power. The proposed method also includes a system identification algorithm, which helps acquire, format, validate, and identify the pattern based on the raw data obtained from a PV system. Autocorrelation and cross-correlation value between input and predicted output 0.02 and 0.15. The model predictive controller helps to recognize the future response of the corresponding PV plant over a specific prediction horizon. The error variation of the predicted values from the actual values for the proposed system is 0.8. The performance analysis of the developed model is compared with the former existing techniques, and the role and aptness of the proposed system in smart grid digitization is also discussed.

Anisotropic constitutive modeling with the considerations of two strategies in describing the distorted hardening of AA6061 under tension-dominated loading conditions

This paper investigates the anisotropic yielding and distorted hardening behaviors of AA6061-T6. Yoon 2014 model was used to describe the anisotropic plastic behaviors of material and two strategies (i.e. different equivalent plastic strain intervals at two levels) in describing the distorted hardening were used. At the integration point level, a program utilizing the forward Euler method and regression mapping algorithms has been developed. This program accepts strain as input and provides corresponding stress as output. Results found that the prediction ability of constitutive model strongly depend on the strain interval of parameters identification. Excessive interval might result in the unexpected fluctuations of stress-strain curve. Such fluctuations primarily come from the deviation of identified parameters and interpolation data. The former deviation is found caused by the changes in initial value during identifying and the latter mainly because the interpolation data departures from the actual values. In addition, a detailed analysis about parameters effects is also conducted for in-depth understanding of parameters identification.

Deep learning model for dynamic color design of all-dielectric metasurfaces.

Silicon nanostructure colors have rapidly developed in recent years, offering high resolution and a broad color gamut that traditional pigments cannot achieve. The reflected colors of metasurfaces are determined by the geometric structure of the unit cell and the refractive index matching layer parameters. It is evident that the design of specific colors involves numerous parameters, making it challenging to achieve through conventional calculations. Therefore, the tandem network instead of conventional electromagnetic simulation is natural. The forward part of the network incorporates feature cross terms to improve accuracy, enabling high-precision predictions of structural colors based on structural parameters. The average color difference between the predicted and actual color values in the L,a,b color space is 1.38. The network has been proven to accurately predict the refractive index and height of the refractive index matching layer during the dynamic tuning process. Additionally, the issue of the inverse network converging to incorrect solutions was addressed by leveraging the characteristics of the activation function. The results show that the color difference between the colors designed by the inverse network compared to the actual colors in the L,a,b color spaces is only 2.22, which meets the requirements for commercial applications.

Enhancing Afforestation Practices in Hilly Terrain: A Study on Soil Disturbance by Earth Augers Based on the Discrete Element Method

Afforestation operations in hilly regions are both arduous and unsafe. The mechanized afforestation method that takes into account soil and water conservation measures is deemed highly important. This paper examines the operational process and the auger’s mechanism of digging below the ground using the discrete element method (DEM). Using this model, soil disturbance parameters and reaction forces are satisfactorily predicted, exhibiting similar trends to experimental observations. This research also examines the influence of key parameters on soil disturbance and distribution patterns and analyzes the conditions and mechanisms of the formation of fish-scale pits to preserve soil and water. A field experiment of pit digging in woodland is carried out to test the performance of the device. The error rates for the actual and simulated values of the efficiency of conveying soil and the distance of throwing soil on plain terrain and slopes were 12.7% and 8.2%, and 8.6% and 15.7%, respectively. Overall, this research provides a theoretical basis for the innovative exploration, development, and optimized design of earth augers in hilly regions.

RGA: a unified measure of predictive accuracy

AbstractA key point to assess statistical forecasts is the evaluation of their predictive accuracy. Recently, a new measure, called Rank Graduation Accuracy (RGA), based on the concordance between the ranks of the predicted values and the ranks of the actual values of a series of observations to be forecast, was proposed for the assessment of the quality of the predictions. In this paper, we demonstrate that, in a classification perspective, when the response to be predicted is binary, the RGA coincides both with the AUROC and the Wilcoxon-Mann–Whitney statistic, and can be employed to evaluate the accuracy of probability forecasts. When the response to be predicted is real valued, the RGA can still be applied, differently from the AUROC, and similarly to measures such as the RMSE. Differently from the RMSE, the RGA measure evaluates point predictions in terms of their ranks, rather than in terms of their values, improving robustness.

Overproduction of bacterial cellulose from Acetobacter xylinum BPR2001 using food industries wastes.

In this study, a cost-effective complex culture media containing molasses and corn steep liquor (CSL) was developed for the high production of bacterial cellulose (BC) by investigating the effect of four effective factors on BC production at three levels using Taguchi and combined methods. The predicted and actual values of BC production in optimal conditions by Taguchi and combined methods were 8.41 and 14.52g/L, respectively. These results showed that the combined method was more suitable for predicting the optimal conditions in the optimization of BC production, the cost of developed culture medium was around 94% cost of HS medium preparation, molasses was the most effective factor in both experimental design methods, and initial pH adjustment had little impact on BC production. Then, the effect of inoculation conditions containing three factors of inoculation age, ethanol addition time, and agitation rate on the increase of BC production at three levels was investigated using the response surface methodology with the Box-Behnken design algorithm. Under the optimal conditions including inoculum age of 3 days, ethanol addition time of 10 days, and stirring speed of 100rpm, the predicted and experimental results of BC production were 21.61 and 20.21g/L, respectively. This is among the highest ever reported for BC production, which was achieved with a more cost-effective culture medium containing molasses and CSL.

A Method for Tomato Plant Stem and Leaf Segmentation and Phenotypic Extraction Based on Skeleton Extraction and Supervoxel Clustering

To address the current problem of the difficulty of extracting the phenotypic parameters of tomato plants in a non-destructive and accurate way, we proposed a method of stem and leaf segmentation and phenotypic extraction of tomato plants based on skeleton extraction and supervoxel clustering. To carry out growth and cultivation experiments on tomato plants in a solar greenhouse, we obtained multi-view image sequences of the tomato plants to construct three-dimensional models of the plant. We used Laplace’s skeleton extraction algorithm to extract the skeleton of the point cloud after removing the noise points using a multi-filtering algorithm, and, based on the plant skeleton, searched for the highest point path, height constraints, and radius constraints to separate the stem from the leaf. At the same time, a supervoxel segmentation method based on Euclidean distance was used to segment each leaf. We extracted a total of six phenotypic parameters of the plant: height, stem diameter, leaf angle, leaf length, leaf width and leaf area, using the segmented organs, which are important for the phenotype. The results showed that the average accuracy, average recall and average F1 scores of the stem and leaf segmentation were 0.88, 0.80 and 0.84, and the segmentation indexes were better than the other four segmentation algorithms; the coefficients of determination between the measurement values of the phenotypic parameters and the real values were 0.97, 0.84, 0.88, 0.94, 0.92 and 0.93; and the root-mean-square errors were 2.17 cm, 0.346 cm, 5.65°, 3.18 cm, 2.99 cm and 8.79 cm2. The measurement values of the proposed method had a strong correlation with the actual values, which could satisfy the requirements of daily production and provide technical support for the extraction of high-throughput phenotypic parameters of tomato plants in solar greenhouses.

HDTO-DeepAR: A novel hybrid approach to forecast surface water quality indicators

Water is a vital resource supporting a broad spectrum of ecosystems and human activities. The quality of river water has declined in recent years due to the discharge of hazardous materials and toxins. Deep learning and machine learning have gained significant attention for analysing time-series data. However, these methods often suffer from high complexity and significant forecasting errors, primarily due to non-linear datasets and hyperparameter settings. To address these challenges, we have developed an innovative HDTO-DeepAR approach for predicting water quality indicators. This proposed approach is compared with standalone algorithms, including DeepAR, BiLSTM, GRU and XGBoost, using performance metrics such as MAE, MSE, MAPE, and NSE. The NSE of the hybrid approach ranges between 0.8 to 0.96. Given the value’s proximity to 1, the model appears to be efficient. The PICP values (ranging from 95% to 98%) indicate that the model is highly reliable in forecasting water quality indicators. Experimental results reveal a close resemblance between the model’s predictions and actual values, providing valuable insights for predicting future trends. The comparative study shows that the suggested model surpasses all existing, well-known models.

Mapping Short Warwick and Edinburgh Mental Wellbeing Scale (SWEMWBS) to Recovering Quality of Life (ReQoL) to estimate health utilities

BackgroundThe Short Warwick and Edinburgh Mental Wellbeing Scale (SWEMWBS) is a widely used non-preference-based measure of mental health in the UK. The primary aim of this paper is to construct an algorithm to translate the SWEMWBS scores to utilities using the Recovering Quality of Life Utility Index (ReQoL-UI) measure.MethodsService users experiencing mental health difficulties were recruited in two separate cross-sectional studies in the UK. The following direct mapping functions were used: Ordinary Least Square, Tobit, Generalised Linear Models. Indirect (response) mapping was performed using seemingly unrelated ordered probit to predict responses to each of the ReQoL-UI items and subsequently to predict using UK tariffs of the ReQoL-UI from SWEMWBS. The performance of all models was assessed by the mean absolute errors, root mean square errors between the predicted and observed utilities and graphical representations across the SWEMWBS score range.ResultsAnalyses were based on 2573 respondents who had complete data on the ReQoL-UI items, SWEMWBS items, age and sex. The direct mapping methods predicted ReQoL-UI scores across the range of SWEMWBS scores reasonably well. Very little differences were found among the three regression specifications in terms of model fit and visual inspection when comparing modelled and actual utility values across the score range of the SWEMWBS. However, when running simulations to consider uncertainty, it is clear that response mapping is superior.ConclusionsThis study presents mapping algorithms from SWEMWBS to ReQoL as an alternative way to generate utilities from SWEMWBS. The algorithm from the indirect mapping is recommended to predict utilities from the SWEMWBS.

Estimation of Power System Inertia with the Integration of Converter-Interfaced Generation via MEMD during a Large Disturbance

The decrease in overall inertia in power systems due to the shift from synchronous generator production to renewable energy sources (RESs) presents a significant challenge. This transition affects the system’s stable frequency response, making it highly sensitive to imbalances between production and consumption, particularly during large disturbances. To address this issue, this paper introduces a novel approach using Multivariate Empirical Mode Decomposition (MEMD) for the accurate estimation of power system inertia. This approach involves applying MEMD, a complex signal processing technique, to power system frequency signals. The study utilizes PMU (Phasor Measurement Unit) data and simulated disturbances in the IEEE 39 bus test system to conduct this analysis. MEMD offers substantial advantages in analyzing multivariate data and frequency signals during disturbances, providing accurate estimations of system inertia. This approach enhances the understanding of power system dynamics in the context of renewable energy integration. However, the complexity of this methodology and the requirement for precise data collection are challenges that need to be addressed. The results from this approach show high accuracy in estimating the rate of change of frequency (RoCoF) and system inertia, with minimal deviation from actual values. The findings highlight the significant impact of renewable energy integration on system inertia and emphasize the necessity of accurate inertia estimation in modern power systems.

Analyzing daily change patterns of indoor temperature in district heating systems: A clustering and regression approach

Measuring the indoor temperature of building rooms is a valuable approach for evaluating thermal comfort and providing feedback control for heat substations in district heating systems (DHSs) in China. Previous studies on indoor temperatures have primarily focused on analyzing their overall trends and influencing factors, while research on daily change patterns is lacking. This study utilized a clustering method to analyze the indoor temperature data from an actual DHS in Northeast China. First, a 24-h observation vector was constructed using the deviation between the actual and target values to represent the daily temperature pattern. Second, the k-means method was applied to cluster the values, and the quantity distribution and typical characteristics of each cluster were analyzed. Finally, a multi-nominal logistic regression model was used to analyze the influence of different factors on each cluster. The comparison results with the four representative clustering algorithms indicated that k-means was the optimal model and the optimal number of clusters was 4. The trend of each cluster was roughly the same, with the main difference being the fluctuation amplitude and distance from the target value. The differences between the clusters were related to various influencing features, with the primary return pressure for workdays and the secondary return pressure for holidays being the most significant. This study identified the optimal daily variation patterns of indoor temperature and analyzed the important features that affect this pattern, which is beneficial for enhancing the regulatory efficiency of DHS.

Urban land use efficiency and its driving factors in ecologically fragile areas: the case of northwest China

ABSTRACT Urban land serves as the medium for urban economic and residential activity. The degree of intensive and efficient use of urban land and the coordinated growth of the economy and society are significantly impacted by utilization efficiency. Northwest China’s temperature, geography, water, and soil characteristics make less of the region suited for urban development. Thus, it is imperative to investigate Northwest China’s urban land use efficiency (ULUE). This study applies the efficiency loss model to highlight the discrepancy between the optimal and actual values of each input-output variable of the ULUE, and uses the Super-SBM model to compute the ULUE of 39 cities. Additionally, the internal driving elements of the ULUE are examined using the Global Malmquist-Luenberger (GML) index. This study discovered that Northwest China had a low ULUE, with a modest upward trend seen between 2004 and 2018. The overall shape of the region was similar to an ‘N’ curve, with high efficiency values concentrated in the southeast of the region in 2018. When considering the agglomeration impact of urban agglomeration, these results are dependable. Moreover, the efficiency loss model discovered that pollution and wasteful resource usage were the main reasons behind the drop in ULUE in Northwest China. Finally, the GML index and its constituents revealed that the main force behind ULUE’s growth in Northwest China is technological developments. This research provides important policy implications for reducing the gap between eastern and western regions, supporting high-quality development in northwest China, and protecting and intensively using urban land.

Toward efficient and accurate extraction of instantaneous frequency with chirplet transform and its applications in bearing fault diagnosis

In many engineering applications, the instantaneous frequency (IF) needs to be identified rapidly and precisely, especially in rotating machinery condition monitoring, where quasi-real-time analysis is a favorable option, and any error in shaft speed (identified from the instantaneous frequency) needs to be reduced to prevent it from accumulating with time. The modern time–frequency analysis (TFA) method provides a way for IF identification; however, sometimes, it still suffers from energy smearing and closely spaced frequency interference. In addition, many TFA methods based on the chirplet transform also present the problem of excessive computation when searching for the optimal parameters. To overcome these limitations, we propose the multi-extracting velocity-synchronous chirplet transform (MEVCT), an enhanced TFA method, in this paper. In the MEVCT, a refined chirp parameter search is developed first to reduce the computational cost while increasing time–frequency accuracy. Then, an iterative high-order IF estimator is constructed to allow the IF to achieve or approach the actual values even in highly nonlinear situations. After that, the obtained IF estimate is substituted into the simultaneous extraction operator to reduce energy dispersion further and eliminate non-reassigned points (NRPs). Simulated and experimental signal analyses demonstrate that MEVCT is able to eliminate NRPs completely. Furthermore, when handling synchronous multicomponent signals, the proposed method has higher IF accuracy with less computational cost and better anti-noise capability.

Enhancing Wave Energy Conversion Efficiency through Supervised Regression Machine Learning Models

The incorporation of machine learning (ML) has yielded substantial benefits in detecting nonlinear patterns across a wide range of applications, including offshore engineering. Existing ML works, specifically supervised regression models, have not undergone exhaustive scrutiny, and there are no potential or concurrent models for improving the performance of wave energy converter (WEC) devices. This study employs supervised regression ML models, including multi-layer perceptron, support vector regression, and XGBoost, to optimize the geometric aspects of an asymmetric WEC inspired by Salter’s duck, based on key parameters. These important parameters, the ballast weight and its position, vary along a guided line within the available geometric resilience of the asymmetric WEC. Each supervised regression ML model was fine-tuned through hyperparameter optimization using Grid cross-validation. When evaluating the performance of each ML model, it became evident that the tuned hyperparameters of XGBoost led to predictions that strongly aligned with the actual values compared to other models. Furthermore, the study extended to assess the performance of the optimized WEC at the designated deployment test site location.

Predicting radiographic outcomes of vertebral body tethering in adolescent idiopathic scoliosis patients using machine learning.

Anterior Vertebral Body Tethering (AVBT) is a growing alternative treatment for adolescent idiopathic scoliosis (AIS), offering an option besides spinal fusion. While AVBT aims to correct spinal deformity through growth correction, its outcomes have been mixed. To improve surgical outcomes, this study aimed to develop a machine learning-based tool to predict short- and midterm spinal curve correction in AIS patients who underwent AVBT surgery, using the most predictive clinical, radiographic, and surgical parameters. After institutional review board approval and based on inclusion criteria, 91 AIS patients who underwent AVBT surgery were selected from the Shriners Hospitals for Children, Philadelphia. For all patients, longitudinal standing (PA or AP, and lateral) and side bending spinal Radiographs were retrospectively obtained at six visits: preop and first standing, one year, two years, five years postop, and at the most recent follow-up. Demographic, radiographic, and surgical features associated with curve correction were collected. The sequential backward feature selection method was used to eliminate correlated features and to provide a rank-ordered list of the most predictive features of the AVBT correction. A Gradient Boosting Regressor (GBR) model was trained and tested using the selected features to predict the final correction of the curve in AIS patients. Eleven most predictive features were identified. The GBR model predicted the final Cobb angle with an average error of 6.3 ± 5.6 degrees. The model also provided a prediction interval, where 84% of the actual values were within the 90% prediction interval. A list of the most predictive features for AVBT curve correction was provided. The GBR model, trained on these features, predicted the final curve magnitude with a clinically acceptable margin of error. This model can be used as a clinical tool to plan AVBT surgical parameters and improve outcomes.

Cost modelling for dimension stone quarry operations

Dimension stone mining is Increasing because of growing demand from the building and construction industries. Diamond wire saw cutting is a popular dimension stone mining technique around the world. Estimation of the production cost for this method is critical for these operations. In this research a production cost model was developed for the mining of granite using a diamond wire saw. Cost and production data were collected from a granitic dolerite operation in Pakistan for a 4.5-year period, and analysed to identify the cost components. Fuel (34%), labour (15%), consumables (40%), and maintenance (5%) were found to be the four major cost components in this operation. A regression-based dimension stone production cost model incorporating statistically significant variables was developed. The data presented a strong correlation of production cost with fuel, labour, and consumables costs. Finally, a production cost model based on the consumption cost of diamond wire is presented. The model was verified using actual production cost data, which indicated 95% (R2 = 0.95) variance in the data between the predicted and actual production cost values. The model may help field engineers in computing their dimension stone operational cost during the cutting phase.

Flow fields prediction for data-driven model of parallel twin cylinders based on POD-RBFNN and POD-BPNN surrogate models

Flow around cylinders is an important phenomenon in many different engineering fields. In this paper, the fast prediction of the pressure fields of parallel twin cylinders is implemented based on a data-driven algorithm. Firstly, the pressure fields of parallel twin cylinders with a low Reynolds number are obtained through the Computational Fluid Dynamics (CFD) method. The pressure fields at different time steps are collected to form a snapshot matrix. The Proper Orthogonal Decomposition (POD) algorithm is then applied to obtain the POD basis vectors of the snapshot matrix, enabling the reconstruing the pressure fields. Subsequently, two reduced-order models (ROM) called the POD-RBFNN and POD-BPNN surrogate models are proposed in this paper. The POD-RBFNN surrogate model uses the Radial Basis Function Neural Network (RBFNN) to train the POD mode coefficients obtained from the POD algorithm, while the POD-BPNN surrogate model uses the Backpropagation Neural Network (BPNN) for the same purpose. Linearly combining the POD mode coefficients predicted by the POD-RBFNN or POD-BPNN surrogate models with the POD basis vectors obtained from the POD algorithm enables fast and efficient prediction of pressure fields for non-sample points. Finally, comparisons are made between the predicted pressure fields obtained from these two surrogate models and the actual values obtained through CFD simulations. It is found that both the POD-RBFNN and POD-BPNN surrogate models proposed in this paper not only significantly improve efficiency but also maintain a high level of accuracy. However, the training time of the POD-RBFNN surrogate model is significantly shorter than that of the POD-BPNN surrogate model. Additionally, the POD-RBFNN surrogate model exhibits smaller Root Mean Square Errors (RMSE) and Mean Absolute Error (MAE). For the data-driven model of parallel twin cylinders described in this paper, the POD-RBFNN surrogate model is more suitable to predict the pressure fields. The research results in this paper are believed to hold significant value for CFD calculations of parallel twin cylinder models, offering essential guidance for a deeper understanding of cylinder flow problems and optimizing engineering design.

Land subsidence analysis along high-speed railway based on EEMD-Prophet method

Environmental changes and ground subsidence along railway lines are serious concerns during high-speed railway operations. It is worth noting that AutoRegressive Integrated Moving Average (ARMA), Long Short-Term Memory (LSTM), and other prediction methods may present limitations when applied to predict InSAR time series results. To address this issue, this study proposes a prediction method that decomposes the nonlinear settlement time series of feature points obtained through InSAR technology using Ensemble Empirical Mode Decomposition (EEMD). Subsequently, multiple Intrinsic Mode Functions (IMFs) are generated, and each IMF is individually predicted using the Prophet forecasting model. Finally, we employ an equal-weight superimposition method to combine the results, resulting in the prediction of the InSAR settlement time series. The predicted values of each component are subsequently weighted equally and combined to derive the final prediction outcome. This paper selects InSAR monitoring data along a high-speed railway in inland China and uses the proposed method and ARMA and Prophet models to carry out comparative experiments. The experimental results show that compared with the ARMA and Prophet models, the method in this paper improves the root mean square error by 58.01% and 32.3%, and increases the mean absolute error by 62.69% and 33.78%, respectively. The predicted settlement values generated by our method exhibit better agreement with the actual InSAR monitoring values.