BP Neural Network Training Research Articles

Thickness Characterization of Steel Plate Coating Materials with Terahertz Time-Domain Reflection Spectroscopy Based on BP Neural Network.

Accurate monitoring of steel plate coating thickness is crucial in construction quality control and durability assessments. To address this challenge, this study introduces a terahertz time-domain reflection spectroscopy based on a BP neural network model to achieve a quantitative visualization characterization of coating thickness. The BP neural network eliminates the inherent dependence of terahertz reflection spectroscopy on the refractive index value in thickness calculation. This trained BP neural network model effectively establishes a functional relationship between signal feature parameters and the corresponding thickness values. Additionally, the proposed model can innovatively measure different coating materials' refractive indexes, revealing the corresponding values for the black paint, white paint, epoxy resin, and rubber as 2.212, 1.967, 1.924, and 2.185, respectively. The experimental results demonstrate the trained BP neural network model possesses remarkable accuracy in predicting coating thickness within the scanning area, achieving a precision level exceeding 96%. This method enables the visualization of coating thickness and the extraction of thickness characterization values. Furthermore, using the thickness imaging results as a reference, the method can accurately identify the thickness abnormalities across the scanning area, locating the position and size of potential defects such as internal scratches and foreign object defects. This innovative approach offers a superior means of monitoring and assessing the thickness distribution quality of the steel plate coating layer materials.

Social entertainment robot based on neural network algorithm in personalized music course simulation

This study focuses on the simulation of personalized music courses using social entertainment robots based on neural network algorithms, and introduces the current research status of music processing algorithms based on neural networks. A detailed explanation was provided on the training methods for neural network algorithms, including BP neural network training and music generation algorithms. Research was conducted on activation functions, music output detection, and loss function detection. A virtual music course system framework has been designed to provide students with personalized learning experiences. In a social entertainment robot based on neural network algorithms, a complete control node system was reconstructed, aiming to achieve interactive and entertainment effects of music courses through sensing technology. The control node system includes multiple key components, and the robot is able to perceive and recognize the actions, facial expressions, and sounds of students. Social entertainment robots have intelligent classification and recommendation functions, and this algorithm can intelligently classify and recommend music resources based on students’ music preferences, levels, and needs. By analyzing the personal preferences and learning progress of students, robots can provide customized music course content and learning materials for each student, improving learning effectiveness and personal interests.

Design of Non-Intrusive Online Monitoring System for Traction Elevators

With the increase in elevator usage, more and more elevator real-time monitoring equipment is being applied to the operation of elevators. Traditional elevator monitoring equipment adopts a multi-sensor decentralized installation and layout, and the monitoring accuracy is low, which directly affects the effective alarm of the monitoring system; however, existing online monitoring systems cannot quickly alarm for faults. Aiming to solve the above problems, an elevator online monitoring system based on narrow-band Internet of Things (NB-IoT) is designed. The system is highly integrated with an STM32 main control chip, a six-axis acceleration gyroscope sensor, and an air pressure sensor to realize the edge calculation of the monitoring system. At the same time, this paper eliminates the temperature drift of the pressure sensor by using a temperature compensation algorithm and inputs the extracted characteristic parameters into the BP neural network for training to eliminate the zero drift so as to obtain the real-time height data of the elevator. The six-axis acceleration gyroscope sensor is used to calculate the posture so as to avoid the problem that a three-axis acceleration sensor or a three-axis gyroscope sensor alone cannot obtain accurate posture data. In order to further improve the monitoring accuracy, the peak-to-peak value of the signal is calculated by using a 95% confidence interval algorithm to reduce the suppression of the high-frequency components of the signal by noise and ensure that the signal has a large signal-to-noise ratio so that the obtained elevator car posture and vibration operation data are more accurate. Finally, the effectiveness of the proposed method is verified by experiments.

Research on the Application of Machine Learning Algorithms in Green Economy Indicator Modeling and Sustainable Development Evaluation

This study presents a comprehensive modeling and analysis framework for investigating the impact of green GDP on climate factors and its potential for promoting sustainable development. Our model utilizes the entropy weight method (EWM) and grey correlation analysis (GRA) to optimize and process data, while employing a BP neural network for model training. Results indicate that green GDP leads to smaller global climate damage compared to traditional GDP, indicating its potential to alleviate the global climate crisis and promote green development.Furthermore, we built a green GDP prediction model using LSTM to forecast future changes. From the perspective of economy and management, the results indicate that transitioning from GDP to green GDP initially leads to a decrease in the index, followed by an increase. This suggests that green GDP development has short-term drawbacks but long-term benefits, including reduced resource consumption, improved economic conditions, and contribution to sustainable development. Finally, we analyzed the learning rate, number of iterations, and thresholds of related parameters through Logistic binary classification. Results demonstrate that our model is relatively stable, highlighting how green GDP addresses ecological challenges while promoting economic prosperity. These findings underscore the urgency of adopting the concept of green GDP in the contemporary era for achieving sustainable development goals.Overall, this research provides critical insights on the potential benefits and feasibility of green GDP as a crucial tool for mitigating climate risks and enhancing global prosperity.

Fault diagnosis of power electronic circuits using optimized BP neural networks

You should leave 8 mm of space above the abstract and 10 mm after the abstract. The heading Abstract should be typed in bold 9-point Arial. The body of the abstract should be typed in normal 9-point Times in a single paragraph, immediately following the heading. The text should be set to 1 line spacing. The abstract should be centred across the page, indented 17 mm from the left and right page margins and justified. It should not normally exceed 200 wordsThree-phase rectifiers have a wide range of applications in industrial production and daily life, and failure to diagnose their faults promptly may affect the reliability of the system operation, resulting in huge safety hazards and economic losses. Therefore, it is of great significance to conduct online fault diagnosis research on the power electronic circuits within three-phase rectifiers. An optimized BP neural network algorithm is proposed for diagnosing open-circuit faults of thyristors in three-phase rectifier circuits. The output voltage waveform characteristics of the circuit, when a fault occurs, are analyzed, and the corresponding output voltage peaks at the same cycle time when different tubes are damaged under the fifth fault type are used as fault feature vectors, and the fault information is input into the BP, optimized BP neural network for training, and the trained neural network is used for fault diagnosis. The fault diagnosis accuracy was obtained by comparing the network output with the desired output. The results of simulation experiments show that the optimized BP neural network can diagnose and analyze the faults of rectifier circuits more efficiently than directly applying the BP network for fault diagnosis.

Construction risk analysis of cantilever casting mid-span closure of bridge in reservoir area

In order to assess the impact of frequent microseismic and cyclic water flow forces on the bridge cantilever casting construction in the Three Gorges reservoir area, a limited element-quantum-orthogonal experimental design-neural network-Monte Carlo simulation method was used to analyze the risk assessment of the construction. The orthogonal experimental design method is utilized to generate probability distribution samples of water velocity, water height, and microseismic acceleration. These samples are then combined with a finite element model and trained using a BP neural network. This approach aims to establish a nonlinear mapping relationship between each control parameter and the axial deviation value in the mid-span closure, and then the Monte Carlo simulation is used to generate the random influence parameter, which can be substituted into the trained BP neural network to predict the risk of the error in the mid-axial deviation. The results suggested that the risk probability of the medial axis deviation of the closure section due to the coupling effect of water flow force and high-frequency microseismicity had reached 5.06%, and the risk of line deviation of the cantilever casting construction of the bridge in the reservoir area needed to be paid sufficient attention and prevented. The risk of line deviation of the cantilever casting construction of the reservoir bridge needs to attract enough attention and precaution.

Diversified Teaching Strategies for College Swimming Courses in the Context of Artificial Intelligence

Abstract Swimming is a sport with a very significant exercise effect, so swimming courses in colleges and universities can promote the overall development of students’ physical fitness. In this paper, students’ swimming data are collected using intelligent sensors, and the collected data are noise-reduced and normalized by a low-pass filter. After the completion of data preprocessing, the swimming posture data features are extracted. The features are dimensionality reduced by the PCA method, combined with the BP neural network for training and accurate swimming posture recognition, and on this basis, a diversified teaching system for swimming courses in smart colleges and universities is constructed. After the students’ swimming data set was collected, the recognition effect of each swimming stroke was analyzed, and a comparison experiment was set up to investigate the practical effect and strategy of this teaching mode. The results show that the average accuracy of recognizing each stroke in this study can reach more than 0.988, and in the teaching test, the average difference between the performance of the experimental group and the control group is 7.6, and the P-value of the technical evaluation performance of the two groups of students in this project is 0.002<0.05, which significantly improves the swimming performance. This study can be combined with artificial intelligence technology to diversify the teaching of swimming courses in colleges and universities.

Dynamic Allocation Mechanism of Cloud Computing Resources Driven by Neural Network

With the popularization of cloud computing technology, the dynamic allocation mechanism of cloud computing resources has become an important research field to improve resource utilization and meet the needs of diversified workloads. The purpose of this study is to explore the dynamic allocation mechanism of cloud computing resources driven by neural network and introduce the powerful ability of deep learning into cloud computing environment. We put forward a comprehensive framework, which combines data collection, analysis, decision-making and implementation to realize intelligent resource allocation. These data will be used to train BP neural network (BPNN). In order to predict the bidding price, a BPNN is designed, which usually includes input layer, hidden layer and output layer. The number of nodes in the input layer is equal to the dimension of the input feature, and the number of nodes in the output layer is 1, which indicates the prediction of the bidding price. Through experiments and simulations, we verify the effectiveness of the dynamic resource allocation mechanism driven by neural network. The results show that this mechanism can better adapt to the changing workload requirements, improve resource utilization and reduce resource waste. In addition, it provides better performance and user experience, thus enhancing the competitiveness of cloud computing systems.

Prediction and evaluation of polymer turbulent drag reduction efficiency based on BP neural network

In the process of oil exploitation and transportation, in order to effectively predict and control energy consumption for drag reduction of oil flow, in this paper a BP neural network was proposed based method for predicting and evaluating the turbulent drag reduction efficiency of polymers, which can greatly improve the current situation of relying on empirical formulas and low generality in polymer turbulent drag reduction efficiency prediction. Based on the experimental data sets of four commercial polymer drag-reducing agents FLOXL, М-Flowtreat, Necadd-447, and FLO MXA, obtained at different polymer concentrations, viscosity, density, and Reynolds number, a BP neural network has been established and the optimal number of neurons in the hidden layer was selected using the root mean square error (RMSE) value to obtain the optimal BP neural network prediction model. The BP neural network prediction models for the four polymer drag-reducing agents all have a good fit of 0.98 or above, and the R2 of the trained BP neural network for the Necadd-447 drag-reducing agents is 0.9949, which is the best among the four polymer drag-reducing agents. The BP neural network established in this paper can be applied to the turbulent drag reduction transport of long-distance pipelines for oil products to achieve the prediction of the drag reduction efficiency of polymer additives.

Parameter calibration method of clustered-particle logic concrete DEM model using BP neural network-particle swarm optimisation algorithm (BP-PSO) inversion method

A high precision back propagation neural network-particle swarm optimization (BP-PSO) algorithm inversion method is proposed to calibrate the microparameters of the clustered-particle logic concrete discrete element method model. The calibration targets include initial damage, failure mode, and mechanical properties of the material. The research utilise 243 training datasets generated through orthogonal experimental design and conducted simulations to train the BP neural network. In addition, a parameter sensitivity analysis is employed on the trained BP neural network to quantify the impact level of each microparameter and guide future macroparameter fine-tuning. The results indicate that the mean absolute percentage error of the BP-PSO inversion method is only 3.79%. This research also study on the concrete failure mechanism by using a mesoscale model based on clustered-particle logic, which considered concrete as a three-phase composite composed of mortar matrix, aggregates, and interfacial transition zone. The crack initiation, propagation and coalescence of DEM model show a good agreement with the experimental results.

Research on the Site Selection Model of Express Service Outlets based on BP Neural Network

With the continuous development of the times, urban planning for the development of the logistics industry puts forward higher requirements, but also for the development of the express industry provides a more potential market, and how to efficiently carry out the site selection of express service outlets has become a difficult problem, so for the current market demand and the current situation. This paper combines the fuzzy comprehensive evaluation method, the establishment of a four-level comprehensive assessment system, the use of BP neural network training, the city express service outlets for scientific site selection and validation analysis, to get the region of the optimal location of each courier service outlets for the city courier service outlets to provide a scientific site selection program.

Optimization of laser spiral welding using Response surface methodology and genetic algorithms

In the laser spiral welding (LSW) process, the welding parameters have a significant impact on the weld quality. In this paper, experiments were conducted and experimental data were collected on galvanized steel sheets using the LSW process, and mathematical models were developed using response surface methodology (RSM) and genetic algorithm (GA) to verify the specific effects of each process parameter on the weld and to perform process optimization. Laser power, welding speed, gap and focal length were selected as the influencing factors, and melt depth, melt width and concave as the output results. In the RSM model we found that the laser power was positively correlated with the weld depth and width, while the welding speed was inversely correlated with the weld depth and width, the gap was positively correlated with the amount of concave, and the focal length had no significant effect on the weld. In the GA model we use a large amount of experimental data for BP neural network training and iterative optimization using a genetic algorithm. Validation experiments were conducted on two models, and the results indicated that the two models had higher accuracy in predicting the welding depth and width compared to predicting the concave. The GA model had an 8% increase in tensile strength and a 25% increase in plasticity of the weld joint obtained from the optimal process compared to the RSM model. The GA model has higher accuracy in optimizing the LSW process.

Application of power electronic circuit design based on BP artificial neural network in environmental monitoring

Abstract Our natural resources have played an important role in building our economy. In addition to the number one energy resource, oil, electric energy is also the best quality and much in demand by society in the history of human civilization. It is because electrical energy can be fully developed and applied by society that the construction of human spiritual civilization has entered industrialization and high-tech information technology. The transmission and application of electric energy has been recognized by society, but how can we make it more reasonable and efficient for easy utilization. In order to solve the problem of how complex power electronic circuit devices can be more widely used with improved technology. This paper establishes the error estimation model of intelligent circuit from data collection and data prediction preprocessing, then proposes to optimize the number of nodes in the hidden layer by using particle swarm optimization algorithm to address the limitations of the traditional BP neural network, and uses the optimized number of nodes in the hidden layer to build a BP neural network structure to train the training sample data, and calculates the error data of the intelligent circuit based on the trained BP neural network to test the sample data. The method established in this paper is used to perform the evaluation of intelligent circuit operation errors. The simulation example shows that the established model can effectively evaluate the smart circuit operation error, and the evaluation accuracy is significantly improved compared with the traditional evaluation method.

Photovoltaic power generation and charging load prediction research of integrated photovoltaic storage and charging station

Aiming at the obvious randomness and intermittent problems of photovoltaic power generation output and charging load of photovoltaic storage and charging station, a photovoltaic power generation prediction model based on long short-term memory neural network and a charging load prediction model based on BP neural network are proposed respectively. In the photovoltaic power generation prediction model, the Pearson correlation coefficient is used to analyze the influencing factors of photovoltaic power generation power, and the factors with high correlation are used as the input variables of the long short-term memory neural network model. Then the data is divided into four types of seasons by K-means cluster analysis, and the processed data sets are put into the prediction model for training. In the charging load prediction model, the processed historical load data is put into the time series-based BP neural network for iterative training, meanwhile the weights and thresholds of the model are adaptively optimized. Finally, the simulation results verify the extremely high prediction accuracy of the proposed prediction model. The results also show that compared with PSO-BP, GA-BP neural network, this charging load prediction model has higher precision and stability.

Glass Classification and Identification based on K-means++ Clustering Analysis and BP Neural Network Method

The weathering caused by the burial environment will cause the ancient glass to exchange material with the environment, and the weathering layer produced by the material exchange will have an impact on the class judgment of ancient glass. In this paper, we analyze the classification rules of high potassium glass and lead-barium glass; select the appropriate chemical composition for subclass division and establish the classification model of glass. After this, a binary classification (high potassium/lead-barium) model is established by BP neural network training, and the subclassified main components of each type of glass derived from the classification model are used as adjustment variables to analyze the chemical composition of unknown categories of glass artifacts and identify the types to which they belong by controlling a single variable.

A new method for dynamic predicting porosity and permeability of low permeability and tight reservoir under effective overburden pressure based on BP neural network

Under the influence of effective overburden pressure, dynamic prediction of porosity and permeability is very important for the formulation and dynamic adjustment of low permeability and tight reservoir development plan. However, the high heterogeneity and complex influencing factors of reservoirs bring great difficulties to the prediction. Firstly, the variation law of porosity and permeability with the increase of effective overburden pressure is determined by testing. On this foundation, data preprocessing is carried out, a sample database is established, and the main controlling factors of geology, fluid and lithology are identified by grey correlation. Finally, based on BP neural network, porosity and permeability prediction training is carried out, and prediction model is established. Considering the prediction error and whether the model is over or under fitting, the results show that the prediction error of the model established by 10 hidden layer neurons and trainbr training function is the best, and the prediction error of porosity and permeability are 3.22% and 8.67% respectively. The model is applied to the field data, and the prediction error of porosity and permeability are 4.18% and 15.85% respectively, which are in line with the oil production law.

Applications of fractional gradient descent method with adaptive momentum in BP neural networks

A novel fractional gradient descent method with adaptive momentum is presented in this paper to improve the convergence speed and stability for BP neural network training. The fractional Grünwald-Letnikov derivative is used for the fractional gradient. The coefficient of the momentum term is set as an adaptive variable, depending on the fractional gradient of the current step and the weight change of the previous step. We give a detailed convergence proof of the proposed method. Experiments on MNIST data sets and XOR problem demonstrate that the fractional gradient descent method with adaptive momentum term can effectively improve convergence speed, maintain stability of BP neural network training, help escape from local minimum points, and enlarge the selection range of the learning rate.

Glass classification and identification based on systematic clustering and BP neural network algorithm

As glass is buried in the ground for years, its surface has been eroded by weathering, thus changing its chemical composition and structure. In this paper, we search for the basis for the classification of high potassium glass and lead-barium glass, and on this basis, we further carry out subclass classification, and finally perform sensitivity analysis on the classification results. After establishing the glass classification model, the artifacts with unknown glass types are predicted, and the data are input into BP neural network for training and prediction to determine their artifact glass types, and sensitivity analysis is performed on the prediction results.

Research on Design Framework of Middle School Teaching Building Based on Performance Optimization and Prediction in the Scheme Design Stage

The good indoor light environment and comfort of the teaching space are very important for students’ physical and mental health. Meanwhile, China advocates energy conservation and emission reduction policies. However, in order to obtain lower building energy consumption, higher thermal comfort, and daylighting, architects use performance simulation software to repeatedly simulate and refine, which is time-consuming and difficult to obtain the best results from three performances. Given this problem, we constructed the design framework in the early stage of the architectural design of the teaching building. In the first stage of the framework, architects optimized the performance objectives of lighting, thermal comfort, and energy consumption, and performed a cluster analysis on the optimized non-dominated solution to provide a reference for the architect. In the second stage of the framework, architects used the data generated in the optimization process to train the BP neural network and use the trained BP neural network to predict the performance of the building. In this paper, we selected Nanjing Donglu Middle School as a case study. The optimization of the building performance was assessed by a genetic algorithm, generating 3000 sets of sample data during the optimization iteration. Then, we analyzed the non-dominated solution of the sample data through the method of cluster analysis and trained the BP neural network with the sample data as a data set. The prediction model with R-values of 0.998 in the training set and test set was obtained by repeatedly debugging the number of neurons in the BP neural network. Finally, five groups of design parameters were randomly selected and brought into the trained BP neural network, and the predictive value was close to the simulated value. The construction of the framework provides design ideas for architects in the early teaching of building design and helps designers to make better decisions.

A neural network-assisted 3D theoretical thermoelastic solution for laminated liquid crystal elastomer plate used in restoring cardiac mechanical function

Some atrial contractile assist devices applied on the heart surface can be regarded as a laminated Liquid crystal elastomer (LCE) plate under steady temperature loads and a contact mechanical force. An exact solution for the deformation of the laminated LCE plate under combined thermal and mechanical loads is derived by solving the three-dimensional (3D) equilibrium equations including heat conduction and thermoelastic theory. The validity of mathematical formula and computer programming is proved by convergence and comparison examples with finite element method (FEM). In order to simplify the complex calculation of exact solution, a back propagation neural network (BPNN) is further trained with a database containing 9504 sets of thermo-mechanical load conditions and their corresponding deformation which is solved by the exact solutions. Then the deformations of LCE plate subject to combined thermo-mechanical load can be predicted by this BP neural network instead of complex numerical calculation. Moreover, it is also applied to inverse the contact mechanical force at the bottom surface of LCE plate with a given deformation and temperature conditions. The results show that: (1) The results from the exact theoretical solution are in consistence with that from FEM but have a higher computational efficiency and stability; (2) The deformation of the laminated plate is more sensitive to the layered thickness of LCE than the variation of the temperature; (3) 3-D elasticity solutions of a laminated LCE plate under the combined thermos-mechanical load can be effectively predicted by a trained BP neural network.