Genetic Algorithm Optimized BP Neural Network Research Articles

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.

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.

Relative density prediction of additively manufactured Inconel 718: a study on genetic algorithm optimized neural network models

PurposeThe quality and properties of Inconel 718 (IN718) from selective laser melting (SLM), a major additive manufacturing (AM) process, have been studied extensively. Among all aspects of quality, relative density (RD) is most widely investigated, and it significantly affects the mechanical properties of SLM-ed materials. This study aims to develop robust RD prediction models based on the data accumulated in literature using machining learning approaches.Design/methodology/approachBy mining the literature of SLM-ed IN718, a comprehensive data set is created, which consists of the four major process parameters of laser power, scan speed, hatch spacing, layer thickness and RD data. A back propagation neural network (BPNN) model, along with its two optimized models: genetic algorithm (GA) optimized BPNN (GA-BPNN) and adaptive GA optimized BPNN (AGA-BPNN) models are created for predicting the RD of SLM-ed IN718, and their prediction performances are compared.FindingsOverall, satisfactory prediction accuracies are obtained – the R2 values of the built BPNN, GA-BPNN and AGA-BPNN models are 73.5%, 75.3% and 79.9%, respectively. This also shows that by incorporating the optimization technique, the prediction accuracy of BPNN is improved and AGA-BPNN has the highest accuracy. Moreover, SLM experiments are conducted to test the model predictability. It is found that the predictions generally agree well with the experiment data, and the order of the model prediction accuracies is consistent with that based on the literature data.Originality/valueThis research highlights that by mining literature data, prediction models of RD of SLM-ed IN718 can be obtained with satisfactory performance, which consider more process parameters and cover wider parameter ranges than any individual studies, in a cost-effective manner.

Process optimization of high-speed dry milling UD-CF/PEEK laminates using GA-BP neural network

High-performance carbon fiber-reinforced polyetheretherketone (CF/PEEK) is widely used in aerospace and premium-end medical fields due to its high strength-weight ratio, shock resistance, and reusability. However, its dry machining requirement is a significant limit to improving machining efficiency and machining quality using a traditional process. Addressing this issue, the high-speed dry (HSD) machining technique is imported in this paper. Multi-level mixed orthogonal experiments of dry milling unidirectional (UD) CF/PEEK laminates with the fiber orientation of 0° and 90° are designed. Aiming at quantitative characterizing surface quality, three-dimensional (3D) surface roughness Sq and 3D fractal dimension Ds are used to present surface roughness and surface defects, respectively. A characterization system for surface defects generated in milling CRF/PEEK is proposed. A prediction model of surface quality considering fiber orientation, cutting speed, feed per tooth, and cutting width is then established using the genetic algorithm optimized BP (GA-BP) neural network. The prediction results show that the model is of acceptable generalization capability with a prediction accuracy of over 90.39%. Based on the analysis of surface qualities and cutting temperatures, the HSD machining technique is verified to be feasible in milling UD-CF/PEEK, and the recommended cutting speed in the HSD milling boundary is 1500–1600 m/min. The 3D fractal dimension is verified feasible to evaluate the size of complex surface defects of the machined UD-CF/PEEK. It has a non-rigid negative correlation with Sq in general. Besides, cutting speed and fiber orientation are the key factors affecting the machined surface microstructural characteristics. The present study gives technical references for improving surface quality in HSD milling CF/PEEK.

Mechanical behavior prediction of CF/PEEK-titanium hybrid laminates considering temperature effect by artificial neural network

This paper focuses on the mechanical behaviors of carbon fiber-reinforced PEEK-titanium hybrid laminates (TiGr) under different temperatures. The tensile strength, flexural strength and inter-laminar shear strength (ILSS) of TiGr laminates with different stacking structures were investigated at room temperature, 80, 150 and 220 °C, respectively. The results show that the tensile strength, flexural strength and ILSS gradually decrease as the temperature increases, and reduced by 12.3%, 45.4% and 29.1% for Type I, 12.9%, 48.8% and 30.1% for Type II, 16.1%, 45.1% and 39.4% for Type III at 220 °C. Moreover, a three-layers back-propagation (BP) neural network was developed to predict the corresponding mechanical behaviors of TiGr laminates, then the genetic algorithm (GA) was introduced to optimize the established BP neural network and the accuracies of the two prediction models were compared. It is found that both the trained BP and GA-BP neural networks could predict the mechanical performances of TiGr laminates well while the GA optimized BP neural network has better prediction results, with average absolute errors of 1.78%, 1.92% and 2.41% for tensile, flexural and inter-laminar shear strengths, respectively. Furthermore, the predictability of the trained models was validated with data from published literature, providing a new concept for the application of BP neural networks to fiber metal laminates.

A prediction model of milling force for aviation 7050 aluminum alloy based on improved RBF neural network

Semi-finishing is an important step to reduce the cutting vibration and deformation caused by the thin-walled stiffening rib parts of aeronautical inertial control products during the finishing process. In the semi-finishing process, the variation of cutting force is a key factor of machining deformation. A large number of samples considering the effect of cutting parameters, such as spindle speed, feed per tooth, and milling width were simulated by the finite element simulation software AdvantEdge to obtain the milling forces based on 7050 aluminum alloy (Al7050) for aviation. Moreover, the standardized Euclidean distance was introduced to the radial basis neural network model (RBF-NN) to develop an improved RBF neural network model (IRBF-NN), and a high-precision model to predict the effect of cutting parameters on the thin-walled semi-finishing milling forces was established based on the proposed IRBF-NN. Results show that the maximum relative errors of milling force obtained by the genetic algorithm–optimized BP neural network model (GA-BP-NN), RBF-NN, and IRBF-NN are 13.9%, 12.5%, and 4.1%, respectively. Accordingly, the proposed IRBF-NN has high accuracy and effectiveness to predict milling force for the semi-finishing of aerospace Al7050 thin-walled stiffening rib parts.

Investigation of Strength and Fatigue Life of Rubber Asphalt Mixture.

Strength and fatigue life are essential parameters of pavement structure design. To accurately determine the pavement structure resistance of rubber asphalt mixture, the strength tests at various temperatures, loading rate, and fatigue tests at different stress levels were conducted in this research. Based on the proposed experiments, the change law of rubber asphalt mixture strength with different temperatures and loading rates was revealed. The phenomenological fatigue equation of rubber asphalt mixture was established. The genetic algorithm optimized backpropagation neural network (GA-BPNN) is highly reliable for optimizing production processes in civil engineering, and it has a remarkable application effect. A GA-BPNN strength and fatigue life prediction model was created in this study. The reliability of the prediction model was verified through experiments. The results showed that the rubber asphalt mixture strength decreases and increases with the increase of temperature and loading rate, respectively. The goodness of fit of the rubber asphalt mixture strength and fatigue life prediction model based on the GA-BPNN could reach 0.989 and 0.998, respectively. The indicators of the fatigue life prediction model are superior to the conventional phenomenological fatigue equation model. The GA-BPNN provides an effective method for predicting the rubber asphalt mixture strength and fatigue life, which significantly improves the accuracy of the resistance design of the rubber asphalt pavement structure.