Genetic Algorithm Optimized BP Neural Network Research Articles

Surface roughness prediction model and surface topography analysis of 2.5D-Cf/SiC in two-dimensional ultrasonic assisted grinding based on GA-BP neural network

Because of the anisotropy and non-uniformity, the grinding surface quality of Cf/SiC composites is difficult to be accurately predicted. The quality of the surface directly determines the assembly accuracy. To better predict the surface quality of 2.5D-Cf/SiC composites, the BP neural network was optimized by genetic algorithm (GA), and the surface roughness prediction model of fiber orientation, ultrasonic amplitude, and machining parameters was established. The empirical formula prediction model of surface roughness was established by the method of multiple linear regression analysis. The results show that the prediction accuracy of GA optimized BP neural network is the best, followed by the empirical formula, and the prediction effect of the BP neural network is the worst. The average absolute percentage error of these three models is 10.045%, 16.912% and 26.6245% respectively. The kinematic models of conventional and two-dimensional ultrasonic-assisted grinding of single particles were established respectively. Based on the kinematic model, the reasons for the reduction of surface roughness by 2D ultrasonic-assisted grinding are explained. According to the orthogonal test results, the surface roughness decreases the most when vs is 0.66m/s, vw is 100mm/min, ap is 150μm along the fiber orientation of 0°, and the maximum percentage of reduction is 53.18%. Increasing the linear speed, reducing the feed speed, reducing the grinding depth, and applying ultrasound can reduce the extrusion pressure along the axis of 0° fiber, and then reduce the length and depth of cracks and thus reduce the surface defects. The axial shear force of the 90° oriented abrasive particles on the fiber is reduced, thus reducing the surface damage caused by torsional deformation. Reduce 3D surface profile and improve surface quality. The maximum percentage reduction of each parameter index is as follows: root mean square height decreased by 28.8%, skew decreased by 88.61%, kurtosis decreased by 48.97%, maximum crest height decreased by 48.55%, maximum sag height decreased by 40.09%, maximum height decreased by 43.93%, and arithmetic mean height decreased by 26.06%.

Monitoring Soybean Soil Moisture Content Based on UAV Multispectral and Thermal-Infrared Remote-Sensing Information Fusion.

By integrating the thermal characteristics from thermal-infrared remote sensing with the physiological and structural information of vegetation revealed by multispectral remote sensing, a more comprehensive assessment of the crop soil-moisture-status response can be achieved. In this study, multispectral and thermal-infrared remote-sensing data, along with soil-moisture-content (SMC) samples (0~20 cm, 20~40 cm, and 40~60 cm soil layers), were collected during the flowering stage of soybean. Data sources included vegetation indices, texture features, texture indices, and thermal-infrared vegetation indices. Spectral parameters with a significant correlation level (p < 0.01) were selected and input into the model as single- and fuse-input variables. Three machine learning methods, eXtreme Gradient Boosting (XGBoost), Random Forest (RF), and Genetic Algorithm-optimized Backpropagation Neural Network (GA-BP), were utilized to construct prediction models for soybean SMC based on the fusion of UAV multispectral and thermal-infrared remote-sensing information. The results indicated that among the single-input variables, the vegetation indices (VIs) derived from multispectral sensors had the optimal accuracy for monitoring SMC in different soil layers under soybean cultivation. The prediction accuracy was the lowest when using single-texture information, while the combination of texture feature values into new texture indices significantly improved the performance of estimating SMC. The fusion of vegetation indices (VIs), texture indices (TIs), and thermal-infrared vegetation indices (TVIs) provided a better prediction of soybean SMC. The optimal prediction model for SMC in different soil layers under soybean cultivation was constructed based on the input combination of VIs + TIs + TVIs, and XGBoost was identified as the preferred method for soybean SMC monitoring and modeling, with its R2 = 0.780, RMSE = 0.437%, and MRE = 1.667% in predicting 0~20 cm SMC. In summary, the fusion of UAV multispectral and thermal-infrared remote-sensing information has good application value in predicting SMC in different soil layers under soybean cultivation. This study can provide technical support for precise management of soybean soil moisture status using the UAV platform.

Prediction of the classification of the pedestrian landing mechanism in pedestrian-vehicle collisions

To predict the classification of the pedestrian landing mechanism in pedestrian-vehicle collisions, 1303 reconstructed real pedestrian-vehicle collision cases were selected, and relevant data from before, during, and after the collisions were extracted. A total of 1303 sets of data with eight parameters were obtained via significance analysis, correlation analysis, and collinearity analysis. Then, the Backpropagation Neural Network (BPNN), Genetic Algorithm (GA) optimized BPNN (GA-BPNN), Principal Component Analysis (PCA) optimized BPNN (PCA-BPNN), Principal Component Analysis (PCA) and Genetic Algorithm (GA) optimized BPNN (PCA-GA-BPNN) were used to construct prediction models for the classification of the pedestrian landing mechanism, and the prediction effects were evaluated. The PCA-GA-BPNN model was found to be the optimal model; the prediction accuracies of the pre-collision, in-collision, and post-collision models were 72.4%, 96.4%, and 96.8%, respectively. Further analysis revealed that the optimal model could also accurately predict the classification of the pedestrian landing mechanism in six cadaver experiments. Additionally, the ratio of the pedestrian height to the vehicle hood height ( R P-V) was found to have an impact on the prediction effect of the model. Thus, an improved model considering R P-V was proposed, and was found to significantly improve the prediction accuracy of pedestrian forward-throwing mechanism. The research results provide new ideas for ground-related injury prediction, and also provide support for pedestrian protection in intelligent vehicles.

Prediction of Geometric Characteristics of Laser Cladding Layer Based on Least Squares Support Vector Regression and Crested Porcupine Optimization.

The morphology size of laser cladding is a crucial parameter that significantly impacts the quality and performance of the cladding layer. This study proposes a predictive model for the cladding morphology size based on the Least Squares Support Vector Regression (LSSVR) and the Crowned Porcupine Optimization (CPO) algorithm. Specifically, the proposed model takes three key parameters as inputs: laser power, scanning speed, and powder feeding rate, with the width and height of the cladding layer as outputs. To further enhance the predictive accuracy of the LSSVR model, a CPO-based optimization strategy is applied to adjust the penalty factor and kernel parameters. Consequently, the CPO-LSSVR model is established and evaluated against the LSSVR model and the Genetic Algorithm-optimized Backpropagation Neural Network (GA-BP) model in terms of relative error metrics. The experimental results demonstrate that the CPO-LSSVR model can achieve a significantly improved relative error of no more than 2.5%, indicating a substantial enhancement in predictive accuracy compared to other methods and showcasing its superior predictive performance. The high accuracy of the CPO-LSSVR model can effectively guide the selection of laser cladding process parameters and thereby enhance the quality and efficiency of the cladding process.

Smart Cushions with Machine Learning-Enhanced Force Sensors for Pressure Injury Risk Assessment.

Prolonged sitting can easily result in pressure injury (PI) for certain people who have had strokes or spinal cord injuries. There are not many methods available for tracking contact surface pressure and shear force to evaluate the PI risk. Here, we propose a smart cushion that uses two-dimensional force sensors (2D-FSs) to measure the pressure and shear force in the buttocks. A machine learning algorithm is then used to compute the shear stresses in the gluteal muscles, which helps to determine the PI risk. The 2D-FS consists of a ferroelectret coaxial sensor (FCS) unit placed atop a ferroelectret film sensor (FFS) unit, allowing it to detect both vertical and horizontal forces simultaneously. To characterize and calibrate, two experimental approaches are applied: one involves simultaneously applying two perpendicular forces, and one involves applying a single force. To separate the two forces, the 2D-FS is decoupled using a deep neural network technique. Multiple FCSs are embedded to form a smart cushion, and a genetic algorithm-optimized backpropagation neural network is proposed and trained to predict the shear strain in the buttocks to prevent PI. By tracking the danger of PI, the smart cushion based on 2D-FSs may be further connected with home-based intelligent care platforms to increase patient equality for spinal cord injury patients and lower the expense of nursing or rehabilitation care.

Application of Genetic Algorithm-Optimized BP Neural Network in Prognosis Prediction of Hemorrhagic Stroke

Clinical intelligent diagnosis and treatment of hemorrhagic stroke, as a combination of artificial intelligence and intelligent medical treatment, is more conducive to perfectly monitoring the pathological cycle changes of patients through data flow. Based on the BP neural network model, this paper constructs a neural network model based on GABP to predict the 90-day mRS Score of patients, and successfully accurately predicts the prognosis of patients with hemorrhagic stroke. Based on the relationship between the patient's prognosis and personal history, disease history, treatment methods and imaging features, the recommendations for clinical decision-making were made. In order to test the distribution of the data, a Shapiro-Wilk distribution test model was constructed to test whether there were significant differences in the distribution of multiple populations, and finally a conclusion was reached that 90-day mRS Was not related to gender. The GABP neural network model constructed in this paper overcomes the limitation of BP neural network in complex problems, and uses genetic algorithm to optimize the weight and threshold, which improves the performance and optimization efficiency of the model. Overall, the GABP-based neural network model represents an approach that integrates neural networks and genetic algorithms and has the potential to solve complex problems. Future research and development will help to further improve the performance, interpretability and applicability of the model, thus promoting its wide application in different fields.

Estimation of Winter Wheat Chlorophyll Content Based on Wavelet Transform and the Optimal Spectral Index

Hyperspectral remote sensing technology plays a vital role in advancing modern precision agriculture due to its non-destructive and efficient nature. To achieve accurate monitoring of winter wheat chlorophyll content, this study utilized 68 sets of chlorophyll content data and hyperspectral measurements collected during the jointing stage of winter wheat over two consecutive years (2019–2020), under various fertilization types and nitrogen application levels. Continuous wavelet transform was applied to transform the original reflectance, ranging from 21 to 210, and the correlation matrix method was utilized to identify the spectral index at each scale, with the highest correlation to winter wheat chlorophyll content as the optimal spectral index combination input. Subsequently, winter wheat chlorophyll content prediction models were developed using three machine learning methods: random forest (RF), support vector machine (SVM), and a genetic algorithm-optimized backpropagation neural network (GA-BP). The results indicate that the spectral data processed through continuous wavelet transform at seven scales, from 21 to 27, show the highest correlation with winter wheat chlorophyll content at a scale of 26, with a correlation coefficient of 0.738, compared with the correlation of 0.611 of the original reflectance, and the accuracy is improved by 20.7%. The average highest correlation value between the spectral index at scale 26 and winter wheat chlorophyll content is 0.752. As the scale of wavelet transform increases, the correlation between the spectral index and winter wheat chlorophyll content and the accuracy of the predictive model show a trend of first increasing and then decreasing. The optimal input variables for predicting winter wheat chlorophyll content and the best machine learning method are the spectral data at a scale of 26 processing combined with the GA-BP model. The optimal predictive model has a validation set coefficient of determination (R2) of 0.859, root mean square error (RMSE) of 1.366, and mean relative error (MRE) of 2.920%. The results show that the prediction model can provide a technical basis for improving the hyperspectral inversion accuracy of winter wheat chlorophyll and modern precision agriculture.

Amphibious vehicle's resistance optimization through neural networks and genetic algorithms

Amphibious vehicles, as a new type of aquatic and terrestrial transport platform, are increasingly involved in the existing transportation system. Resistance is a key factor that affects the efficiency and energy consumption of vehicles in aquatic sailing. Resistance reduction optimization design is a focal point and challenge in the design process of amphibious vehicles. In this paper, a resistance performance optimization method has been proposed based on neural networks and genetic algorithms. First, key parameters for the shape design are extracted based on a thorough understanding of the vehicle's performance. These parameters are used to construct a parameterized design space. Second, a training set is obtained based on the Latin hypercube sampling method and numerical calculation methods, and a test set is randomly generated. To achieve better resistance prediction performance, a method based on the genetic algorithm-optimized backpropagation Neural Network is proposed. Next, the resistance performance of the two operating conditions is optimized through the non-dominated sorting genetic algorithm II, and optimized configuration parameters are obtained, which has a 22.71% energy-saving ratio at cruising speed. Finally, the optimized configuration is analyzed using numerical calculation methods to validate the resistance prediction and optimization methods.

Short term load prediction of regional heating and heat storage system based on neural network

Accurate heat load prediction is the key to achieve fine control, energy conservation, and carbon reduction of regional hydronics. Taking the regional hydronics of a city in the north of China as the research object, the author, respectively uses back propagation neural network (BPNN), genetic algorithm (GA) optimized BPNN (GA-BPNN), and autoregressive integrated moving average model (ARI?MA) combined BPNN (ARIMA BPNN) to predict its heat load, and compares the accuracy and applicability of each prediction method. The results indicate that GA-BPNN has the smallest prediction error, followed by ARIMA-BPNN, but the latter requires less data for prediction. In practical engineering, if there is a sufficient amount of data related to heat load, it is recommended to use GA-BPNN. If there is a small amount of data, ARIMA-BP prediction method can be used.

Investigation on compressive strength of coral aggregate concrete: Hybrid machine learning models and experimental validation

The compressive strength of coral aggregate concrete (CAC-CS) holds importance in structural engineering and architectural design. Thus, this study establishes machine learning models for the prediction of CAC-CS and assesses their accuracy and generalization capability. The importance of input variables is analyzed by using Shapley additive explanations (SHAP) and single-variable fluctuations. A user-friendly graphical user interface (GUI) for CAC-CS prediction is developed to facilitate practical use. The accuracy of the GUI is subsequently validated through experiments. Results demonstrate that the genetic algorithm-optimized backpropagation neural network (GA-BP) model provides predictions that are closely aligned with experimental values. The GA-BP model minimizes the mean and standard deviation of its residual distribution. The performance indicators, including correlation coefficient, mean absolute error, mean absolute percentage error, mean square error, and root mean square error, of the GA-BP model are 0.98, 3.78, 10.95, 11.08, and 3.33, respectively, and are superior to those of the other tested models. The analysis of SHAP values and single-variable fluctuations reveals that among the factors influencing CAC-CS, water-binder ratio and curing age are the most sensitive. Reducing the water-binder ratio and increasing the curing age enhances CAC-CS. Additionally, the sensitivity analysis of mineral admixtures and the water-binder ratio in experiments aligns with the analysis of SHAP values and single-variable fluctuations. The error between the experimental values and GUI predictions for 14 mixture designs is less than 5 %, thus validating the accuracy and generalization capability of the GUI proposed in this study. In conclusion, this research provides practical guidelines for preparing CAC and contributes to the conservation and development of island resources.

State of health estimation for lithium-ion batteries based on current interrupt method and genetic algorithm optimized back propagation neural network

Accurate state-of-health (SOH) estimation is an essential prerequisite for a battery management system (BMS) to improve battery utilization efficiency. The impedance information can be utilized to reflect the SOH. However, the traditional electrochemical impedance spectroscopy (EIS) method suffers from time-consuming measurements and specialized equipment. This study aims to establish a connection between EIS and the current interrupt method, which reduces the difficulty of obtaining impedance information through its utilization. This method can be applied in real time during charging, enabling it to be incorporated into a BMS. A genetic algorithm optimized back propagation neural network (GA-BPNN) is developed to estimate the SOH based on the impedance information obtained in the current interrupt method as inputs. The genetic algorithm improves the weights and thresholds of the neural network, which solves the parameter calibration problem. In this study, besides utilizing measurement data from different aging conditions as training set data, hybrid tests comparable to the actual usage environment are employed as validation set data. The experimental results show that a combination of the proposed current interrupt method and GA-BPNN can estimate SOH accurately with the root mean square error (RMSE) as low as 0.77 % in a complex hybrid test environment.

Prediction of Surface Residual Stresses after Laser Shock Processing on TC4 Titanium Alloy Using Different Neural Network Agent Models

Nowadays, it has become a trend to use finite element simulation instead of experimental processes, and this is widely used in the fields of structural mechanics, fluid mechanics, fracture mechanics, and so on. By replacing the experimental process with finite element simulation, we can reduce time and costs; however, when using finite element simulation, we need to define a series of settings, such as modeling, material assignment, environment settings, and many other operations. For laser shock processing intensification, the simulation experiment process is cumbersome and time-consuming. It involves performing neural network agent modeling, replacing finite element simulation with the learning and prediction capabilities of neural networks, learning by using some of the simulation results as a training sets for the neural network, and then learning by using the remaining simulation results as testing sets to test the predictive ability of the neural network agent model. TC4 titanium alloy was selected as the experimental material. Three kinds of neural network agent models, a genetic algorithm-optimized BP network, a strong classifier design based on BP_Adaboost, and an extreme learning machine, instead of finite element simulation experiments, were used to predict the residual stresses generated on the surfaces of the material under different laser shock parameters. Comparing the prediction performances of different neural network agent models, the genetic algorithm-optimized BP network shows the best prediction performance, and its prediction value matches well with the experimental value. The R2, RMSE, and MAE of the testing sets of the BP network optimized using the genetic algorithm were 0.9985, 44.4518, and 30.6285, respectively. The BP network agent model optimized using the genetic algorithm for laser shock parameters other than the 208 sets of data also had good prediction performance, and the predicted values were similar to the actual experimental results. The prediction results show that the BP network optimized using the genetic algorithm can predict the residual stresses on the surface of TC4 titanium alloy material under strengthening via laser shock processing; the genetic algorithm-optimized BP neural network agent model is more convenient and quicker compared to the finite element simulation, and the predicted value is also similar to the actual value. It can thus be used to replace finite element simulation by establishing a more convenient and quicker neural network agent model.

Prediction of elastic properties of 3D4D rotary braided composites with voids using multi-scale finite element and surrogate models

The conventional evaluation of 3D braided composites' mechanical properties through numerical and experimental methodologies serves as a hindrance to material application owing to the considerable expenses, time constraints, and laborious efforts involved. Moreover, the presence of void defects induced during the processing exacerbates this challenge. In this study, a multi-scale finite element model (FEM) and a surrogate model are established for predicting elastic properties of three dimensional four directional (3D4D) rotary braided composites with voids for the first time. Based on the established FEM, a comprehensive dataset containing 768 data points is formed, covering the ranges of both design parameters and void defect parameters. The influence of braiding angle, yarn width, and porosity, on the elastic constants of 3D4D rotary braided composites is accurately analyzed. A genetic algorithm-optimized back propagation neural network (GABPNN) model is developed, which possess the capability to replicate FEM outcomes with a commendable R-value of 0.99. The remarkable concordance between the anticipated outcomes and experimental datasets corroborates the triumphant implementation of the present method in unraveling the interconnections between microstructure and properties in 3D4D rotary braided composites containing voids. Consequently, this offers a propitious instrument for expediting the intelligent conception and refinement of composite materials.

Spatial Downscaling of Soil Moisture Based on Fusion Methods in Complex Terrains

Large-area soil moisture (SM) data with high resolution and precision are the foundation for the research and application of hydrological and meteorological models, water resource evaluation, agricultural management, and warning of geological disasters. It is still challenging to downscale SM products in complex terrains that require fine spatial details. In this study, SM data from the Soil Moisture Active and Passive (SMAP) satellite were downscaled from 36 to 1 km in the summer and autumn of 2017 in Sichuan Province, China. Genetic-algorithm-optimized backpropagation (GABP) neural network, random forest, and convolutional neural network were applied. A fusion model between SM and longitude, latitude, elevation, slope, aspect, land-cover type, land surface temperature, normalized difference vegetation index, enhanced vegetation index, evapotranspiration, day sequence, and AM/PM was established. After downscaling, the in situ information was fused through a geographical analysis combined with a spatial interpolation to improve the quality of the downscaled SM. The comparative results show that in complex terrains, the GABP neural network better captures the soil moisture variations in both time and space domains. The GDA_Kriging method is able to merge in situ information in the downscaled SM while simultaneously maintaining the dynamic range and spatial details.

Temperature field test and prediction using a GA-BP neural network for CRTS II slab tracks

AbstractThe CRTS II slab track, which is connected in a longitudinal direction, is one of the main ballastless tracks in China, with approximately 7365 km of operational track. Temperature loading is a very vital factor leading to slab track damages such as warping and cracking. While existing research on temperature distribution rests on either site tests in special environments or theoretical analysis, the long-term temperature field characteristics are not clear. Therefore, a long-term temperature field test for the CRTS II slab track on bridge-subgrade transition section was conducted to analyze the temperature field. A GA-BP (genetic algorithm optimized back propagation) neural network was trained on the test data to predict the temperature field. The vertical and lateral temperature distributions in four typical days were carried out. We found that the temperature along the track was distributed in a nonlinear manner. This was particularly distinct in the vertical direction for depths of less than 300 mm. The highest and lowest daily temperatures and the daily range of the temperature were analyzed. With the increasing depth, the daily highest temperatures and range of the temperature were smaller, the daily lowest temperatures were higher, and the time corresponding to this peak value appeared later in the day. Both the highest and lowest daily temperature could be predicted using the GA-BP neural network, though the accuracy in predicting the highest temperature was higher than that in predicting the lowest temperature.

Machine Learning-Based Quantitative Structure-Activity Relationship and ADMET Prediction Models for ERα Activity of Anti-Breast Cancer Drug Candidates

Breast cancer is presently one of the most common malignancies worldwide, with a higher fatality rate. In this study, a quantitative structure-activity relationship (QSAR) model of compound biological activity and ADMET (Absorption, Distribution, Metabolism, Excretion, Toxicity) properties prediction model were performed using estrogen receptor alpha (ERα) antagonist information collected from compound samples. We first utilized grey relation analysis (GRA) in conjunction with the random forest (RF) algorithm to identify the top 20 molecular descriptor variables that have the greatest influence on biological activity, and then we used Spearman correlation analysis to identify 16 independent variables. Second, a QSAR model of the compound were developed based on BP neural network (BPNN), genetic algorithm optimized BP neural network (GA-BPNN), and support vector regression (SVR). The BPNN, the SVR, and the logistic regression (LR) models were then used to identify and predict the ADMET properties of substances, with the prediction impacts of each model compared and assessed. The results reveal that a SVR model was used in QSAR quantitative prediction, and in the classification prediction of ADMET properties: the SVR model predicts the Caco-2 and hERG(human Ether-a-go-go Related Gene) properties, the LR model predicts the cytochrome P450 enzyme 3A4 subtype (CYP3A4) and Micronucleus (MN) properties, and the BPNN model predicts the Human Oral Bioavailability (HOB) properties. Finally, information entropy theory is used to validate the rationality of variable screening, and sensitivity analysis of the model demonstrates that the constructed model has high accuracy and stability, which can be used as a reference for screening probable active compounds and drug discovery.

Simulation of the Ozone Concentration in Three Regions of Xinjiang, China, Using a Genetic Algorithm-Optimized BP Neural Network Model

Accurate ozone concentration simulation can provide a health reference for people’s daily lives. Simulating ozone concentrations is a complex task because near-surface ozone production is determined by a combination of volatile organic compounds (VOCs) and NOx emissions, atmospheric photochemical reactions, and meteorological factors. In this study, we applied a genetic algorithm-optimized back propagation (GA-BP) neural network, multiple linear regression (MLR), BP neural network, random forest (RF) algorithm, and long short-term memory network (LSTM) to model ozone concentrations in three regions of Xinjiang, China (Urumqi, Hotan, and Dushanzi districts) for the first time by inputting wind speed, humidity, visibility, temperature, and wind direction. The results showed that the average relative errors of the model simulations in the Urumqi, Hotan, and Dushanzi districts were BP (61%, 14%, and 16%), MLR (97%, 14%, and 23%), RF (39%, 11%, and 14%), LSTM (50%, 12%, and 16%), and GA-BP (16%, 4%, and 6%) and that the significance coefficients R2 were BP (0.73, 0.65, and 0.83), MLR (0.68, 0.62, and 0.74), RF (0.85, 0.80, and 0.88), LSTM (0.78, 0.74, and 0.85), and GA-BP (0.92, 0.93, and 0.94), respectively, with the simulated values of GA-BP being the closest to the true values. The GA-BP model results showed that among the 100 samples with the same wind speed, humidity, visibility, temperature, and wind direction data, the highest simulated ozone concentrations in the Urumqi, Hotan, and Dushanzi districts were 173.5 μg/m3, 114.3 μg/m3, and 228.4 μg/m3, respectively. The results provide a theoretical basis for the effective control of regional ozone pollution in urban areas (Urumqi), dusty areas (Hotan), and industrial areas (Dushanzi) in Xinjiang.

Genetic algorithm-optimized backpropagation neural network establishes a diagnostic prediction model for diabetic nephropathy: Combined machine learning and experimental validation in mice

Genetic algorithm-optimized backpropagation neural network establishes a diagnostic prediction model for diabetic nephropathy: Combined machine learning and experimental validation in mice

Effect of Construction Errors in Cable Forces of Single-Story Orthogonal Cable Network Structures Based on GA-BPNN

The construction process of cable net structure is complicated, which leads to the strong randomness of construction errors. The safety state of the cable net structure is very sensitive to construction errors. Obtaining the coupling relationship between construction errors and cable force response efficiently and accurately is critical to developing the construction technique of cable structures. This paper proposed an analysis method based on a genetic algorithm optimized back propagation neural network (GA-BPNN) to judge the influence of construction error on the cable force of single-layer orthogonal cable network structures. Taking the speed skating stadium of the 2022 Winter Olympic Games as the research object, this paper analyzed the structure form of the venue. According to the characteristics of cable network structure and GA-BPNN calculation, the principle of construction error analysis was put forward. The influence of construction errors of load-bearing cables and stable cables on cable force response was analyzed. The influence degree of different component errors on structural cable forces was obtained, and the most unfavorable key components were obtained. For the key components, the influence trend of different construction errors on cable force was analyzed, and the fitting formula was given. Driven by GA-BPNN, it can realize the analysis of structural and mechanical responses under the action of multi-type, multi-component, and multi-combination construction errors. The results show that the research method efficiently and accurately obtains the performance law of structural cable force under the influence of construction error, effectively predicts the influencing factors of the structural safety risk, and effectively avoids structural safety accidents caused by construction error. The construction errors analysis method based on GA-BPNN proposed in this paper can provide a reference for similar structural analysis and application.

Generating a Long-Term Spatiotemporally Continuous Melt Pond Fraction Dataset for Arctic Sea Ice Using an Artificial Neural Network and a Statistical-Based Temporal Filter

The melt pond fraction (MPF) is an important geophysical parameter of climate and the surface energy budget, and many MPF datasets have been generated from satellite observations. However, the reliability of these datasets suffers from short temporal spans and data gaps. To improve the temporal span and spatiotemporal continuity, we generated a long-term spatiotemporally continuous MPF dataset for Arctic sea ice, which is called the Northeast Normal University-melt pond fraction (NENU-MPF), from Moderate Resolution Imaging Spectroradiometer (MODIS) data. First, the non-linear relationship between the MODIS reflectance/geometries and the MPF was constructed using a genetic algorithm optimized back-propagation neural network (GA-BPNN) model. Then, the data gaps were filled and smoothed using a statistical-based temporal filter. The results show that the GA-BPNN model can provide accurate estimations of the MPF (R2 = 0.76, root mean square error (RMSE) = 0.05) and that the data gaps can be efficiently filled by the statistical-based temporal filter (RMSE = 0.047; bias = −0.022). The newly generated NENU-MPF dataset is consistent with the validation data and with published MPF datasets. Moreover, it has a longer temporal span and is much more spatiotemporally continuous; thus, it improves our knowledge of the long-term dynamics of the MPF over Arctic sea ice surfaces.