BP Neural Network Model Research Articles

Numerical Simulation and ANN Prediction of Crack Problems within Corrosion Defects.

Buried pipelines are widely used, so it is necessary to analyze and study their fracture characteristics. The locations of corrosion defects on the pipe are more susceptible to fracture under the influence of internal pressure generated during material transportation. In the open literature, a large number of studies have been conducted on the failure pressure or residual strength of corroded pipelines. On this basis, this study conducts a fracture analysis on buried pipelines with corrosion areas under seismic loads. The extended finite element method was used to model and analyze the buried pipeline under seismic load, and it was found that the stress value at the crack tip was maximum when the circumferential angle of the crack was near 5° in the corrosion area. The changes in the stress field at the crack tip in the corrosion zone of the pipeline under different loads were compared. Based on the BP algorithm, a neural network model that can predict the stress field at the pipe crack tip is established. The neural network is trained using numerical model data, and a prediction model with a prediction error of less than 10% is constructed. The crack tip characteristics were further studied using the BP neural network model, and it was determined that the tip stress fluctuation range is between 450 MPa and 500 MPa. The neural network model is optimized based on the GA algorithm, which solves the problem of convergence difficulties and improves the prediction accuracy. According to the prediction results, it is found that when the internal pressure increases, the corrosion depth will significantly affect the crack tip stress field. The maximum error of the optimized neural network is 5.32%. The calculation data of the optimized neural network model were compared with the calculation data of other models, and it was determined that GA-BPNN has better adaptability in this research problem.

Tension anisotropy of rolled AA7075-T6 auminium alloy: Experiments, BP-Neural network modeling and orientation dependent failure mechanisms

This paper presents the investigations on anisotropic tension mechanical responses of AA7075-T6 based on experiments, classical strength theory, BP-neural network modeling and fracture morphology characterization. Results show that the tension strength anisotropy weakens with deformation degree. Compared to the traditional method, the machine learning model exhibits more flexible in solving the anisotropic plastic responses of AA7075-T6 auminium alloy sheet and provides more accurate predictions. Through analyzing the fracture surface of tension specimen at various orientations, the failure mechanism is sensitive to orientation. Specifically, the irregular distribution of dimples zones and cleavage steps can be observed at lower material orientation. As the orientation increases, the alternative occurrence of ductile and brittle features dominates the failure mechanism. The medium-size dimple caused by coalescence of small-size dimples represents a transition between ductile features zone and brittle characteristics region.

Optimization of Power Prediction of BP Network with Improved Pelican Algorithm

Abstract In view of the large fluctuation of photovoltaic output power affected by different weather, accurate prediction of photovoltaic output power is particularly important for the safe and stable operation of power system. Firstly, the pelican optimization algorithm ( POA ) is improved in the following three aspects : adding Circle chaotic map to make the population evenly distributed, introducing mutation factor to expand the search range of pelican when approaching prey, adding adaptive weight and firefly disturbance to avoid falling into local optimum in the water surface flight stage ; then, in order to improve the prediction accuracy of BP algorithm, the improved pelican algorithm ( IPOA ) is used to optimize the weights and thresholds of BP neural network, and the IPOA-BP photovoltaic power prediction model is built to improve the accuracy of power prediction. Finally, this paper tests the prediction performance of IPOA-BP, POA-BP and basic BP power prediction models in sunny, cloudy and rainy days through experiments. The experimental results demonstrate that the IPOA-BP prediction model outperforms both the POA-BP and traditional BP neural network models under various weather conditions.

Effects of Seasonal Freezing and Thawing on Soil Moisture and Salinity in the Farmland Shelterbelt System in the Hetao Irrigation District

Water resources are scarce, and secondary soil salinization is severe in the Hetao Irrigation District. Farmland shelterbelt systems (FSS) play a critical role in regulating soil water and salt dynamics within the irrigation district. However, the understanding of soil water and salt migration within FSS during the freeze–thaw period remains unclear due to the complex and multifaceted interactions between water and salt. This study focused on a typical FSS and conducted comprehensive monitoring of soil moisture, salinity, temperature, and meteorological parameters during the freeze–thaw period. The results revealed consistent trends in air temperature and soil temperature overall. Soil freezing durations exceeded thawing durations, and both decreased with an increasing soil depth. At the three critical freeze–thaw nodes, the soil moisture content at a 0–20 cm depth was significantly lower than at a 40–100 cm depth (p < 0.05). The soil water content increased with time and depth at varying distances from the shelterbelt, with an average increase of 7.63% after freezing and thawing. The surface water content at the forest edge (0.3H, 4H) was lower than inside the farmland (1H, 2H, 3H). Soil salt accumulation occurred during both freezing stable periods and melting–thawing periods in the 0–100 cm soil layer near the forest edge (0.3H, 4H), with the highest soil salinity reaching 0.62 g·kg−1. After the freeze–thaw period, the soil salt content in each layer increased by 11.41–47.26% compared to before the freeze–thaw period. Salt accumulation in farmland soil near the shelterbelt was stronger than in the far shelterbelt. The multivariate statistical model demonstrated goodness of fit for soil water and salt as 0.94 and 0.72, respectively, while the BP neural network model showed goodness of fit for soil water and salt as 0.82 and 0.85, respectively. Our results provide an efficient theoretical basis for FSS construction and agricultural water management practices.

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.

Effect Evaluation of Staged Fracturing and Productivity Prediction of Horizontal Wells in Tight Reservoirs

In this paper, the effect evaluation and production prediction of staged fracturing for horizontal wells in tight reservoirs are studied. Firstly, the basic characteristics and value of horizontal wells in tight reservoirs are introduced, their geological characteristics, flow mechanism and permeability model are analyzed and the application of grey theory in effect analysis is discussed. Considering the problems of staged fracturing effect evaluation and the production prediction of horizontal wells in tight reservoirs, a BP neural network model based on deep learning is proposed. Due to the interference of multiple physical parameters and the complex functional relationship in the development of tight reservoir fracturing, the traditional prediction method has low accuracy and it is difficult to establish an accurate mapping relationship. In this paper, a BP neural network is used to simulate multivariable nonlinear mapping by modifying the model, and its advantages in solving the coupling relationship of complex functions are brought into play. A neural network model with fracturing parameters as input and oil and gas production as output is designed. Through the training and testing of data sets, the accuracy and applicability of the proposed model for effect evaluation and yield prediction are verified. The research results show that the model can fit the complex mapping relationship between fracturing information and production and provide an effective evaluation and prediction tool for the development of the staged fracturing of horizontal wells in tight reservoirs.

Characterizing spatio-temporal variations of dimethyl sulfide in the Yellow and East China Sea based on BP neural network

Dimethyl sulfide (DMS), an organic volatile sulfide produced from Dimethylsulfoniopropionate (DMSP), exerts a significant impact on the global climate change. Utilizing published literature data spanning from 2005 to 2020, a BP neural network (BPNN) model of the surface seawater DMS in the Yellow and East China Sea (YECS) was developed to elucidate the influence of various marine factors on the DMS cycle. Results indicated that the six parameters inputted BPNN model, that include the time (month), latitude and longitude, sea-surface chlorophyll a (Chl-a), sea-surface temperature (SST), and sea-surface salinity (SSS), yielded the optimized simulation results (R2 = 0.71). The optimized estimation of surface seawater DMS in the YECS were proved to be closely aligned with the observed data across all seasons, which demonstrated the model’s robust applicability. DMS concentration in surface seawater were found to be affected by multiple factors such as Chl-a and SST. Comparative analysis of the three environmental parameters revealed that Chl-a exhibited the most significant correlation with surface seawater DMS concentration in the YECS (R2 = 0.20). This underscores the pivotal role of chlorophyll in phytoplankton photosynthesis and DMS production, emphasizing its importance as a non-negligible factor in the study of DMS and its sulfur derivatives. Furthermore, surface seawater DMS concentration in the YECS exhibited positive correlations with Chl-a and SST, while displaying a negative correlation with SSS. The DMS concentration in the YECS show substantial seasonal variations, with the maximum value (5.69 nmol/L) in summer followed in decreasing order by spring (3.96 nmol/L), autumn (3.18 nmol/L), and winter (1.60 nmol/L). In the YECS, there was a gradual decrease of DMS concentration from the nearshore to the offshore, especially with the highest DMS concentration concentrated in the Yangtze River Estuary Basin and the south-central coastal part off the Zhejiang Province. Apart from being largely composed by the release of large amounts of nutrients from anthropogenic activities and changes in ocean temperature, the spatial and temporal variability of DMS may be driven by additional physicochemical parameters.

Introduct Risk Prediction Protection Study Based on GA-BP and K-means Clustering

This paper mainly focuses on the risk management problems caused by frequent extreme weather events. Firstly, a grey bp neural network model is used as a heuristic optimization algorithm to construct an IWM model, which is used to decide whether insurance company provides insurance under extreme weather conditions. Secondly, based on the Comprehensive evaluation method (RSR), a DDI evaluation model was constructed to analyze the data of 100 buildings in 10 provinces in China to help developers evaluate the rationality of decisions. Finally, a PM evaluation model is constructed based on AHP and EWM-CRITIC, and K-means clustering is used to conduct comprehensive rating and grading evaluation of 30 low-insurance landmarks, and corresponding protection strategies are designed. The comprehensive application of these models provides useful suggestions and decision support for insurance company underwriting, real estate bidding, community building protection and so on.

Accurate measurement techniques and prediction approaches for the in-situ rock stress

The precise calculation and evaluation of the in-situ rock stress tensor is a crucial factor in addressing the major challenges related to subsurface engineering applications and earth science research. To improve the accuracy of in-situ stress measurement and prediction, an improved overcoring technique involving a measurement circuit, temperature compensation, and calculation method is presented for accurately measuring the in-situ rock stress tensor. Furthermore, an embedded grey BP neural network (GM–BPNN) model is established for predicting in-situ rock stress values. The results indicate that the improved overcoring technique has significantly improved the stress measurement accuracy, and a large number of valuable stress data obtained from many mines have proved the testing performance of this technique. Moreover, the mean relative errors of the prediction results of GM(0, 1) for the three principal stresses all reach 6–30%, and the accuracy of the model fails to meet the requirements. The average relative errors of the prediction results of the BPNN model are all less than 10%, and the model accuracy meets the requirements and has sufficient credibility. Compared with the GM and BPNN models, the embedded GM–BPNN model produces the best results, with mean relative errors of 0.0001–4.8338%. The embedded GM–BPNN model fully utilizes the characteristics of grey theory and BP neural network, which require a small sample size, weaken the randomness of the original data, and gradually approach the accuracy of the model, making it particularly suitable for situations with limited stress data.

Improved BP Neural Network Model for Predicting Regional Agricultural Irrigation Water

Improved BP Neural Network Model for Predicting Regional Agricultural Irrigation Water

Mechanical performance degradation investigation on FRP reinforced concrete based on neural network design method

To predict the effect of chemical-freezing coupling erosion on the properties of four kinds of FRP-reinforced concrete. Rapid freeze–thaw tests were conducted. The mass loss rate, relative dynamic elastic modulus, compressive strength, and flexural capacity were tested to investigate the Mechanical Performance of specimens. The compression specimens are cylindrical specimens wrapped with FRP, and the flexural specimens are pasted with FRP prismatic specimens on the pre-cracked side. A database was built based on 45 groups of experimental test results, and the prediction effect of the BP neural network and CNN model on compressive strength and flexural capacity was compared, respectively. The results showed that CNN did a better job. Finally, the maximum number of freeze–thaw cycles of different FRP-reinforced specimens was predicted based on the CNN model with mass loss rate and relative dynamic modulus as the evaluation criteria. This method can provide a new perspective for predicting the durability of FRP-reinforced concrete.

Predicting the Reasons of Employee Turnover Based on BP Neural Network Model

Employee turnover is related to the survival and development of enterprise team, which increases the cost of human resources and causes management difficulties and economic losses. Bp neural network model can deal with nonlinear problems, large-scale data and complex pattern recognition tasks, and plays an important role in the field of prediction. In this paper, BP neural network model is used to deeply analyze the causes of employee turnover, and deeply predict the causes of complex employee turnover with economic development and post-epidemic era. Through training and learning of relevant data of employee turnover, BP neural network can extract potential patterns and rules of employee turnover, so as to predict the trend of employee turnover in the future. So that enterprises can better understand the demands and expectations of employees, formulate more reasonable policies and plans, improve the job satisfaction and loyalty of employees, and reduce the employee turnover rate. Provide guidance for enterprise management's decision-making and rectification.

Study on insurance in earthquake-prone areas based on BP neural network and logistic regression model

Study on insurance in earthquake-prone areas based on BP neural network and logistic regression model

Research on Vegetables Sales Profit Based on Machine Learning

The sales volume of different vegetables in supermarkets is different, and the profit is different. The research on the profit of each category of vegetables can help supermarkets predict the sales revenue more accurately to improve inventory management, purchasing plans, and sales strategy. This can help the supermarket avoid excess or insufficient inventory, improve profits, and help the supermarket identify potential sales risks. The research team obtained daily sales data for various categories of vegetables from a supermarket on the Internet for some time. First, according to the "cost plus pricing" method of vegetables, regression analysis was conducted on the total cost and total profit of various categories of vegetables, and it was found that there was a strong linear relationship, and the profit rate of various categories of vegetables was relatively fixed. They were [0.6629, 0.5915, 0.4663, 0.5121, 0.5596, 0.5381]. Since the sales volume of vegetables has seasonal characteristics, this paper established the ARIMA model to forecast the sales volume of each category of vegetables; and obtained the sales volume distribution of each category of vegetables from July 1 to 7, 2023. The BP neural network model with 22 layers of hidden layers was established, and the purchase unit price data could be obtained according to the known sales data. The R square of the model was greater than 0.7, and the goodness of fit was high. According to the known correlation, considering the impact of the loss rate on the profit and the rationality of the data, a multivariate linear programming model is established to give the replenishment volume and pricing strategy from July 1 to 7, 2023. It is concluded that the maximum profit of the supermarket in these seven days is [845.1835, 870.5028,861.6210, 858.8551, 839.4595, 864.7580, 892.9897]. Through the sensitivity analysis of the profit rate of each category of vegetables, it is concluded that the profit rate increases by 1%, the income will increase by about 1.1%, and the model is robust.

Multidisciplinary design and optimization of journal bearing for high-power wind turbine speed increaser

With the increasing power of wind turbine generators (WTG), the failure rate of rolling bearings in wind turbines due to insufficient bearing capacity increases with the increase of bearing size. To overcome these challenges, this paper proposes a design optimization method for the rectangular groove elliptical sliding bearing (RGEB) of the WTG output shaft. This method can maximize the radial bearing capacity under the premise of ensuring that the oil film pressure is large, the end leakage and the temperature rise are small. The multidisciplinary design and modeling of RGEB are carried out according to the structure and working conditions of WTG. Based on computational fluid dynamics (CFD), the influence of the number of rectangular dynamic pressure grooves on the bearing performance is analyzed. The kriging model, BP neural network model, and SSA-BP model are constructed for each performance index of RGEB, to establish a high-precision combined surrogate model based on global error criterion. After establishing the optimization equation, the PSO and SQP combined optimization algorithm is used to optimize the optimal structural parameters. The results show that the bearing capacity of the optimized RGEB is 95.9% higher than that of the ordinary elliptical bearing (EB) without increasing the size and quality of the bearing. In addition, the combined surrogate model can effectively replace the expensive finite element model to deal with the WTG sliding-bearing optimization problem.

Temperature compensation model for non-dispersive infrared CO2 gas sensor based on WOA-BP algorithm

Temperature compensation is the main measure to solve the problem that the detection accuracy of non-dispersive infrared CO2 gas sensor is affected by temperature. As the measurement accuracy of the non-dispersive infrared CO2 gas sensor is easily affected by the ambient temperature, this article analyzes the reasons why the sensor is affected by temperature, and proposes a temperature compensation method that integrates the Whale Algorithm (WOA) and BP neural network. The whale algorithm is used to optimize the weights and thresholds of the BP neural network to build a temperature compensation model for the non-dispersive infrared CO2 gas sensor and compare the superiority with the traditional BP neural network model and particle swarm optimization (PSO) BP neural network model. The experimental results show that the temperature compensation model error of WOA-BP algorithm is lower than 30 ppm, and the average absolute error percentage is 3.86%, which is far better than BP neural network and PSO-BP neural network, and effectively reduces the influence of temperature on the accuracy of the sensor.

Horizontal Tail Flight Load Prediction Method for Large UAVs based on Machine Learning

In today’s rapidly developing large UAVs field, flight safety and reliability are essential considerations. Flight load is one of the important factors affecting the load capacity of large UAVs, so the research on load prediction technology can improve flight safety and the reliability of air transportation. This paper first introduces the collection and processing method of flight load data, then uses the machine learning method-BP neural network to predict flight load, establishes the BP neural network model to predict flight load with flight parameters, and optimizes the model according to error analysis. Finally, the accuracy and practicability of the model are verified by analysing the fitting of the training set and the test set, and the error distribution of the true value of shear, bending moment, torque and the absolute percentage of the predicted value, and the model is evaluated.

Assessing the Severity of Verticillium Wilt in Cotton Fields and Constructing Pesticide Application Prescription Maps Using Unmanned Aerial Vehicle (UAV) Multispectral Images

Cotton Verticillium wilt is a common fungal disease during the growth of cotton, leading to the yellowing of leaves, stem dryness, and root rot, severely affecting the yield and quality of cotton. Current monitoring methods for Verticillium wilt mainly rely on manual inspection and field investigation, which are inefficient and costly, and the methods of applying pesticides in cotton fields are singular, with issues of low pesticide efficiency and uneven application. This study aims to combine UAV remote sensing monitoring of cotton Verticillium wilt with the precision spraying characteristics of agricultural drones, to provide a methodological reference for monitoring and precision application of pesticides for cotton diseases. Taking the cotton fields of Shihezi City, Xinjiang as the research subject, high-resolution multispectral images were collected using drones. Simultaneously, 150 sets of field samples with varying degrees of Verticillium wilt were collected through ground data collection, utilizing data analysis methods such as partial least squares regression (PLSR) and neural network models; additionally, a cotton Verticillium wilt monitoring model based on drone remote sensing images was constructed. The results showed that the estimation accuracy R2 of the PLSR and BP neural network models based on EVI, RENDVI, SAVI, MSAVI, and RDVI vegetation indices were 0.778 and 0.817, respectively, with RMSE of 0.126 and 0.117, respectively. Based on this, an analysis of the condition of the areas to be treated was performed, combining the operational parameters of agricultural drones, resulting in a prescription map for spraying against cotton Verticillium wilt.

Research on the Prediction of Infiltration Depth of Xiashu Loess Slopes Based on Particle Swarm Optimized Back Propagation (PSO-BP) Neural Network

The Xiashu loess exhibits expansion when in contact with water and contraction when water is lost, making it highly susceptible to the influence of rainfall. Therefore, it is essential to investigate the infiltration behavior of rainwater in Xiashu loess slopes under various conditions. The depth of infiltration in slopes directly affects the depth of landslide failure and serves as an important indicator for studying slope infiltration characteristics; only a handful of academics have delved into its study. This article is based on on-site rainfall experiments on Xiashu loess slopes, using three main factors, rainfall intensity, rainfall duration, and slope angle, as discrimination indicators for the infiltration depth of Xiashu loess slopes. The particle swarm optimization algorithm is employed to optimize the BP neural network and establish a PSO-BP neural network prediction model. The experimental data are accurately predicted and compared with the multivariate nonlinear regression model and traditional BP neural network models. The results demonstrate that the PSO-BP neural network model exhibits a better fit and higher prediction accuracy than the other two models. This model provides a novel approach for rapidly determining the infiltration depth of Xiashu loess slopes under different rainfall conditions. The results of this study lay the foundation for the prediction of the landslide damage depth and infiltration of Xiashu loess slopes.

BP neural network model for reverse design of two-dimensional triangular lattice photonic crystals

By learning the related properties of photonic crystals, two-dimensional photonic crystals with triangular lattice are selected for study. The point defects are set up and the photon band calculation is carried out by using MPB. The calculated sample data is imported into MATLAB, and the relationship between point defects and photonic energy bands is found by building Bp neural network model, which provides a new method to solve the reverse design of photonic crystals.