Optimized Back Propagation Neural Network Research Articles

Study on the Quality Analysis and Improvement of Tennis Teaching under the Internet System

Improving the quality of tennis teaching and learning is an important part of the modernization of education in China. The limitations of traditional tennis teaching quality evaluation methods have made them controversial, and it is crucial to improve the scientific, rational, and timely evaluation of tenn is teaching quality. Therefore, it is necessary to establish a scientific and rational quality evaluation model for tennis education to evaluate the quality of tennis education. Based on the principle of constructing a perfect tennis teaching quality evaluation system, this paper analyzes the advantages and disadvantages of the previous tennis teaching quality evaluation methods and summarizes the problems existing in the current tennis teaching quality evaluation system in a university. On this basis, to break the limitations of the existing tennis teaching quality evaluation system in a university, a tennis teaching quality evaluation model based on genetic algorithm (GA) and back-propagation (BP) neural network is proposed and a more scientific reasonable tennis teaching quality evaluation system. The evaluation results of BP neural network optimization method based on adaptive mutation genetic algorithm are very satisfactory.

Back-propagation neural network modeling for a pulse tube refrigerator with passive displacer

• The experiment is conducted on a pulse tube refrigerator with passive displacer. • The back-propagation neural network is developed to simulate the performance. • The predicted results of the performance agree with the experimental results. • A relative Carnot efficiency of 23.2% is obtained at 140 K in the prediction. The pulse tube refrigerator (PTR) with passive displacer is a potential alternative for high-efficiency PTR. One of the main problems is that massive experiments are expensive and time-consuming. A back-propagation neural network (BPNN) model is proposed to overcome this challenge. The cooling capacity and relative Carnot efficiency of the PTR with passive displacer can be predicted instead of experiments when the charging pressure, operating frequency, input electrical power, and cooling temperature are determined. The optimal BPNN with the least mean square error is obtained by optimizing the network structure, transfer function, and training function. As a result, the coefficient of determination (R 2 ) values for the training and testing data are 0.9995 and 0.9992, respectively. The predicted values are achieved within eight milliseconds, agreeing with the experimental values. The maximum relative Carnot efficiency of 23.2% occurs at 140 K with a cooling capacity of 15 W. It is concluded that the BPNN is an effective model for simulating the performance of the PTR with passive displacer quickly. Furthermore, the mathematical formulas and the capability to predict the effects of the operating parameters are demonstrated, which is scarce in the literature on the PTR. The BPNN model can be applied to accelerate the development of the PTR with passive displacer.

BPNN Based Indoor Fingerprinting Localization Algorithm Against Environmental Fluctuations

For mobile users, the ability to accurately acquire their own location is critical. By locating, mobile users can get information about their environment and access location-related services. RSS fingerprint-based indoor localization method collects a training database of measurement fingerprints and uses a machine learning classifier to determine a person’s location from a new fingerprint. However, as the environment changes over time due to furniture or other objects being moved, the new fingerprints diverge from those in the original database. Therefore, an RSS-difference based localization system is designed to deal with the above problem. This method combines back-propagation neural network (BPNN) and weighted K-Nearest Neighbor (WKNN) method to improve the fingerprint similarity based indoor location method (FSIL). We train BPNN in the off-line stage to obtain the optimal BPNN parameter settings. In the online stage, K nearest neighbor points are firstly selected based on the improved FSIL algorithm, and then the difference of signal strength values between the K nearest neighbor points and the target user is input into the BPNN network, to obtain the Euclidean distance between the K nearest neighbor points and the target user, and finally the WKNN algorithm is used to obtain the user’s ultimate location. Simulation experiments based on the LDPL model and the Wireless Insite software, as well as the test results based on the indoor localization dataset IPIN2016, show that the localization accuracy in complex indoor scenarios can be improved by at least 11% when using the method proposed in this paper.

A combined model based on sparrow search optimized BP neural network and Markov chain for precipitation prediction in Zhengzhou City, China

Abstract Simulation and prediction of precipitation time series changes are important for revealing global climate change patterns and understanding surface hydrological processes. However, precipitation is influenced by a variety of factors together, showing the characteristics of nonlinear variation patterns. Given that backpropagation (BP) neural network has a strong mapping ability for nonlinear fitting, we consider using BP neural network for precipitation prediction, then use Sparrow Search Algorithm (SSA) to optimize BP network initial threshold and weight information to improve the efficiency of precipitation prediction. To further enhance model predictive performance, the Markov model is employed to predict the residual series of the SSA-BP model, so as to finally construct a combined SSA-BP-Markov model of precipitation. In this paper, the model is used to simulate the rainfall prediction in Zhengzhou City, Henan Province, China, and to compare and analyze with the other traditional models. The empirical prediction results show that the SSA-BP-Markov model is more accurate and the convergence of the algorithm is better. The model provides a new way of thinking for precipitation prediction and is also useful for predicting precipitation in other regions.

Prediction of EPB Shield Tunneling Advance Rate in Mixed Ground Condition Using Optimized BPNN Model

Tunneling in mixed ground often results in severe torque fluctuations and a low advance rate. Therefore, choosing a reasonable set of parameters for accurate advance rate prediction is paramount to reduce cutter wear and improve tunneling efficiency. However, since the geological parameters in mixed ground conditions are diverse and uncertain, the prediction of the advance rate (AR) of EPB shield tunneling is significantly more difficult than that in homogeneous ground (i.e., full-face hard-rock ground). In addition, the operating parameters of the EPB shield tunneling can be subjective and suboptimal, and each of them has some intricate influence on AR. In this paper, an optimized back-propagation neural network by genetic algorithm (BPNN-GA) was proposed for reasonable operating parameter selection and accurate AR prediction, and four typical machine learning methods were used for comparison. Five processing strategies with different input parameters were also proposed and compared to determine the optimum selection of geological parameters in mixed ground conditions. The proposed models with strategies were adopted in the case study of the Nanjing Metro Line S6 project, and a total of 1188 rings of datasets were used for this study. The results showed that the proposed modified BPNN with the genetic algorithm could be effectively implemented for the AR prediction. It concluded that Strategy B—i.e., using the composite ratio and the geological parameters of each layer as input—was the best strategy in mixed ground conditions for advance rate prediction. Hence, a high correlation between measured and predicted AR was observed in this study with a correlation coefficient (R2) of 0.920.

Prediction of GPS Satellite Clock Offset Based on an Improved Particle Swarm Algorithm Optimized BP Neural Network

Satellite clock offset is an important factor affecting the accuracy of real-time precise point positioning (RT-PPP). Due to missing real-time service (RTS) products provided by the International GNSS Service (IGS) or network faults, users may not obtain effective real-time corrections, resulting in the unavailability of RT-PPP. Considering this issue, an improved back propagation (BP) neural network optimized by heterogeneous comprehensive learning and dynamic multi-swarm particle swarm optimizer (HPSO-BP) is proposed for clock offset prediction. The new model uses the particle swarm optimizer to optimize the initial parameters of the BP neural network, which can avoid the instability and over-fitting problems of the traditional BP neural network. IGS RTS product data is selected for the experimental analysis; the results demonstrate that the average prediction precision of the HPSO-BP model for 20-min and 60-min is better than 0.15 ns, improving by approximately 85% compared to traditional models including the linear polynomial (LP) model, the quadratic polynomial (QP) model, the gray system model (GM (1,1)), and the ARMA time series model. It indicates that the HPSO-BP model has reasonable practicability and stability in the short-term satellite clock offset prediction, and its prediction performance is superior to traditional models. Therefore, in practical applications, the clock offset products predicted by the HPSO-BP model can meet the centimeter-level positioning accuracy requirements of RT-PPP.

Research on detection and recognition methods of gas pipelines based on acoustic signal feature analysis

In the process of reconstruction and expansion of gas pipeline, it is easy to destroy in-service gas pipeline and cause safety accidents. In order to realize the detection of in-service pipelines, based on the characteristics of low sound pressure level and easy attenuation of acoustic signals of gas pipelines, the detection and identification method of gas pipelines based on acoustic signal feature analysis was studied by using Hilebert–Huang transform algorithm and optimized Back Propagation (BP) neural network. This method takes the gas pipeline flow noise signals obtained by numerical simulation and experimental verification as the research object, and the underwater acoustic signals are collected for comparative analysis. Empirical Mode Decomposition (EMD) was used to decompose the two signals, and the time-domain waveform of Intrinsic Mode Functions (IMF) component was obtained, and the characteristic parameters of peak value and peak frequency were determined. The energy characteristic parameters of Hilbert marginal spectrum were calculated, and the characteristic database of gas pipeline flow noise signal was obtained. The optimized BP neural network was used for pattern recognition. The results show that the identification rate of gas pipeline acoustic signal can reach 97.5% by using this method, which verifies the effectiveness of the gas pipeline detection and identification method in this paper.

New Anomaly Network Intrusion Detection System in Cloud Environment Based on Optimized Back Propagation Neural Network Using Improved Genetic Algorithm

Cloud computing is distributed architecture, providing computing facilities and storage resource as a service over an open environment (Internet), this lead to different matters related to the security and privacy in cloud computing. Thus, defending network accessible Cloud resources and services from various threats and attacks is of great concern. To address this issue, it is essential to create an efficient and effective Network Intrusion System (NIDS) to detect both outsider and insider intruders with high detection precision in the cloud environment. NIDS has become popular as an important component of the network security infrastructure, which detects malicious activities by monitoring network traffic. In this work, we propose to optimize a very popular soft computing tool widely used for intrusion detection namely, Back Propagation Neural Network (BPNN) using an Improved Genetic Algorithm (IGA). Genetic Algorithm (GA) is improved through optimization strategies, namely Parallel Processing and Fitness Value Hashing, which reduce execution time, convergence time and save processing power. Since, Learning rate and Momentum term are among the most relevant parameters that impact the performance of BPNN classifier, we have employed IGA to find the optimal or near-optimal values of these two parameters which ensure high detection rate, high accuracy and low false alarm rate. The CloudSim simulator 4.0 and DARPA’s KDD cup datasets 1999 are used for simulation. From the detailed performance analysis, it is clear that the proposed system called “ANIDS BPNN-IGA” (Anomaly NIDS based on BPNN and IGA) outperforms several state-of-art methods and it is more suitable for network anomaly detection.

Optimisasi Backpropagation Neural Network dalam Memprediksi IHSG

Covid-19 has become a global epidemic and has spread to many countries in the world, including Indonesia. The COVID-19 pandemic is one source of uncertainty that causes financial data to fluctuate and cause data to be volatile. This outbreak had an impact on financial data, not only on the Rupiah exchange rate but also on the Jakarta Composite Index (JCI). The uncertainty of the JCI makes it difficult for investors, data managers, and business people to predict data for the future. JCI is one indicator of the capital market (stock exchange). The uncertainty of the JCI data causes the need for predictions, so that investors, data managers, and business people can make the right decisions so that they can reduce risk and optimize profits when investing. One of the factors causing the JCI's decline was sentiment caused by investor panic over the rapid spread of COVID-19 in various cities in Indonesia. This research uses Backpropagation Neural Network (BPNN) in making predictions and continues with optimization of BPNN using ensemble techniques. Historical data from the JCI used were obtained from yahoo.finance. The ensemble technique used consists of two approaches, namely combining different architectures and initial weights with the same data and combining different architectures and weights. The results of network performance using ensemble technique optimization show good performance and can outperform the individual network performance of BPNN. Keywords: prediction, JCI, Optimization, BPNN, volatile

Cutting-state identification of machine tools based on improved Dempster-Shafer evidence theory

The reliability of machine tools is highly influenced by the cutting state. The traditional recognition method of cutting state is emphasized on a single classifier, which has the weakness of low identification accuracy and strong randomness. This paper proposes a cutting-state identification method based on improved Dempster-Shafer (DS) evidence theory. This method is divided into multi-classifier preliminary-diagnosis layer and improved DS information-fusion layer. The wavelet packet analysis method is extracted as the input of multi-classifier (Back Propagation (BP) neural network, genetic algorithm (GA) optimized BP neural network and thinking evolution (mind evolutionary algorithms) MEA optimized BP neural network). After the preliminary judgment, the improved DS information-fusion method is integrated as the final judgment, and finally, the effectiveness and feasibility of the method are verified.

Wind Farm Energy Storage System Based on Cat Swarm Optimization–Backpropagation Neural Network Wind Power Prediction

To solve the instability problem of wind turbine power output, the wind power was predicted, and a wind power prediction algorithm optimized by the backpropagation neural network based on the CSO (cat swarm optimization) algorithm was studied, and a wind farm energy storage system model was built on this basis. By collecting the wind power plant’s historical wind speed, power, and other parameters, the short-term wind farm output power was predicted, and the operation of the wind farm energy storage system was controlled to suppress the output power of the wind farm when the wind farm was connected to the grid so as to improve the stability of the output power of the wind farm. At the same time, typical wind farm data were taken as an example to verify the feasibility of the proposed method.

Structural Design and Parameter Optimization of Bionic Exhaust Tailpipe of Tractors

The exhaust tailpipe of a certain type of tractor was improved from the perspective of bionics, and bionic triangular convex texture was added to the inner surface of the exhaust tailpipe. The bionic tailpipe was proposed to improve noise reduction performance without changing the overall size parameters of the prototype tailpipe. Acoustics simulation software was used to predict the aeroacoustics noise and transmission loss of the exhaust tailpipe. Bionic exhaust tailpipes with triangular textures of different numbers of circumferential columns, height, and top angles were analyzed to study the noise reduction performance. The results showed that the proposed bionic exhaust tailpipes with triangular convex textures reduced the total sound pressure level and improved the transmission loss of the prototype exhaust tailpipe. To increase the transmission loss, a genetic algorithms (GA) optimized back-propagation neural network (BP) was used to optimize the bionic triangular convex texture parameters. By studying the aerodynamic noise reduction mechanism of bionic tailpipes, the research suggested that a secondary vortex appeared near the bionic texture and reduced aerodynamic drag and aeroacoustics noise. In addition, the sound pressure level amplitude nephogram, velocity vector nephogram, and velocity amplitude nephogram of the exhaust tailpipes were analyzed to study the vibration noise reduction mechanism of the bionic tailpipes. Then, the noise reduction performance was experimentally evaluated. The experimental results of the bionics exhaust tailpipes with triangular convex textures were analyzed and compared to that of the prototype tailpipe. The results demonstrated that the bionic exhaust tailpipes were able to attenuate noise.

Backpropagation Neural Network optimization and software defect estimation modelling using a hybrid Salp Swarm optimizer-based Simulated Annealing Algorithm

Software Defect Estimation (SDE) is a fundamental problem solving mechanism in the field of software engineering (SE). SDE is a task that identifies software models that are likely to have defects. In addition, SDE plays a vital overall role in improving software quality, reducing software development costs and accelerating software development processes. The Backpropagation Neural Network (BPNN) is a popular machine learning (ML) estimator widely utilized in SE estimation problems. Unfortunately, its performance depends on the initial weight and bias values. Metaheuristic optimization algorithms, as an alternative method, have proven to have strengths in parameter optimizations. Additionally, population-based metaheuristic algorithms suffer from low exploitation capabilities. In this paper, a new hybrid metaheuristic algorithm-based BPNN (SSA–SA) is proposed by hybridizing the Salp Swarm Algorithm (SSA) with the Simulated Annealing (SA) algorithm. The main goal of the hybridization is to adjust the balance between exploration and exploitation in SSA. The proposed algorithm is also assembled with the BPNN estimator to optimize its parameters to reduce the overall estimation error, which boosts the estimation accuracy. Thus, the proposed algorithm addresses the SDE problem. Experimental results prove the superiority of the proposed hybrid algorithm in optimizing BPNN parameters in comparisons against other estimators and algorithms in most SDE datasets and evaluation criteria.

CONSTRUCTION OF SWITCH NONLINEAR DYNAMIC SYSTEM USING ADAPTIVE NEURAL NETWORK TECHNOLOGY

In order to improve the intelligent tracking efficiency of switched nonlinear dynamic system, first, the optimized backpropagation neural network (BPNN) is proposed and the switched nonlinear dynamic system is constructed. Then, the optimized BPNN is extended and the adaptive control is studied. Finally, the extended neural network system related to the error is formed, and then the simulation analysis is carried out. The results show that under different reference signal conditions, the output peak values of the system are 1.7 and 1.5 at 2[Formula: see text]s, respectively, while the response peak value of the reference signal is 1. However, after 4[Formula: see text]s, the peak values of the two signals are very close; under different reference signals, the scaling factor decreases with the increase of time. After 6[Formula: see text]s, the value of scaling factor is 0. Under the condition of reference signal 1, the estimated values of parameters decrease with the increase of time, and tend to 0 after 6[Formula: see text]s. However, under the condition of reference signal 2, the estimated values of parameters in the first 6[Formula: see text]s increase with the increase of time; the peak values of the state variable [Formula: see text] and the adaptive parameters in the system are in a stable state, while the peak values of the system controller and the system tracking error gradually decrease with the increase of time, and finally tend to a stable value. It suggests that the technique proposed in this work can guarantee the signal in switched nonlinear dynamic system to be bounded.

Colorimetric characterization of color imaging systems using a multi‐inputPSO‐BPneural network

AbstractMost commonly used camera characterization methods do not use a deep learning‐based artificial neural network approach at present. This article proposes a colorimetric characterization method for color imaging systems based on the multi‐input particle swarm optimization backpropagation neural network. Combined with a particle swarm optimization algorithm for global search and a 19‐input vector, this method not only overcomes the effects of local extrema on the multi‐input backpropagation neural network, but also improves the accuracy of the common input backpropagation neural network. Images of a ColorChecker SG chart were collected using a Canon EOS 1000D camera for experimental verification, and the color differences were used to evaluate the characterization results. The results show that the color differences of the multi‐input particle swarm optimization backpropagation neural network (structure: 19‐7‐3) model are substantially better than those of the multi‐input backpropagation neural network (structure: 19‐7‐3) and common input backpropagation neural network (structure: 3‐4‐3) models. Its performance is close to that of the weighted nonlinear regression model. The multi‐input particle swarm optimization backpropagation neural network is hence an effective method for colorimetric characterization with good prediction accuracy.

Real-Time Aeromagnetic Compensation With Compressed and Accelerated Neural Networks

As neural networks become an increasingly popular technique in the field of aeromagnetic compensation, there is an increasing demand for hardware systems with more computing power. Compared with the linear regression method, applying a neural network to the task of real-time compensation is difficult because of insufficient computing resources in the unmanned aerial vehicle (UAV) flight detection platform. To perform real-time compensation calculations with limited computing resources, we optimized back propagation neural network (OBPNN) through model compression and acceleration. In this study, we found that the most time-consuming part of network training is the iterative updating of the weights in the BPNN interference model. Using transfer learning, we replace the randomly initialized weights (RWs) with pretrained weights, thereby greatly reducing the number of iterations required. We also apply other model compression and acceleration algorithms. In a case study of our new technique, we implement the fast training of the OBPNN on a Raspberry Pi 4B system. This network processes approximately 316 samples per 0.1 s, which is fast enough to complete aeromagnetic compensation in real time.

Improved Particle Swarm Optimization-Based BP Neural Networks for Aero-Optical Imaging Deviation Prediction

In this article, an improved particle swarm optimization (IPSO) algorithm based on similarity and random mutation is raised. The diversity of particles in the population is decided by the size of the aggregation. When the aggregation degree of particles in the population surpass a certain threshold, the concept of similarity is used to measure the similarity between particles and global extremum, and the particles with higher similarity are discretized by mutation strategy. By increasing the particle swarm’s diversity, the population’s local and global search ability tend to balance. The weight and threshold of the back propagation (BP) neural networks are optimized by the IPSO algorithm. Then, the model of the improved particle swarm optimization back propagation neural network (IPSO-BP) is applied to the aero-optical imaging deviation prediction. The results show that the prediction accuracy of the IPSO-BP model is superior to the PSO-BP model, the extreme learning machine (ELM) model, and the least square support vector machine (LSSVM) model, and its convergence speed is faster than that of the PSO-BP neural network model. Finally, the application of deep learning in aero-optical imaging deviation prediction is analyzed compared with the IPSO-BP neural network model.

Individual thermal comfort models based on optimized BP neural network algorithms

Thermal comfort plays an important role in human life and it affects occupant satisfaction, health, and productivity. Individual differences are not considered in traditional control strategies based on temperature setpoints. The reality is that operators often expend more energy to maintain the indoor environment and the thermal satisfaction of occupancy is not as well as expected. Thus, individual thermal comfort models based on physiological parameters and environmental parameters were presented using the back-propagation (BP) neural network. Moreover, we used three training algorithms including Levenberg-Marquardt (L-M), Bayesian Regularization, and Scaled Conjugate. We observed that using the L-M algorithm resulted in slightly better performance (R=0.96) than other algorithms. The precision results suggest that the BP network algorithm is an effective approach for real-time predicting thermal comfort. In the follow-up study, we would focus on feature engineering (feature selection) and introduce appropriate variables (e.g., heart rate) to improve the model’s accuracy.

Multi-Feature Optimization Study of Soil Total Nitrogen Content Detection Based on Thermal Cracking and Artificial Olfactory System

To improve the accuracy of detecting soil total nitrogen (STN) content by an artificial olfactory system, this paper proposes a multi-feature optimization method for soil total nitrogen content based on an artificial olfactory system. Ten different metal–oxide semiconductor gas sensors were selected to form a sensor array to collect soil gas and generate response curves. Additionally, six features such as the response area, maximum value, average differential coefficient, standard deviation value, average value, and 15th-second transient value of each sensor response curve were extracted to construct an artificial olfactory feature space (10 × 6). Moreover, the relationship between feature space and soil total nitrogen content was used to establish backpropagation neural network (BPNN), extreme learning machine (ELM), and partial least squares regression (PLSR) models were used, and the coefficient of determination (R2), root mean square error (RMSE), and the ratio of performance to deviation (RPD) were selected as prediction performance indicators. The Monte Carlo cross-validation (MCCV) and K-means improved leave-one-out cross-validation (K-means LOOCV) were adopted to identify and remove abnormal samples in the feature space and establish the BPNN model, respectively. There were significant improvements before and after comparing the two rejection methods, among which the MCCV rejection method was superior, where values for R2, RMSE, and RPD were 0.75671, 0.33517, and 1.7938, respectively. After removing the abnormal samples, the soil samples were then subjected to feature-optimized dimensionality reduction using principal component analysis (PCA) and genetic algorithm-based optimization backpropagation neural network (GA-BP). The test results showed that after feature optimization the model indicators performed better than those of the unoptimized model, and the PLSR model with GA-BP for feature optimization had the best prediction effect, with an R2 value of 0.93848, RPD value of 3.5666, and RMSE value of 0.16857 in the test set. R2 and RPD values improved by 14.01% and 50.60%, respectively, compared with those before optimization, and RMSE value decreased by 45.16%, which effectively improved the accuracy of the artificial olfactory system in detecting soil total nitrogen content and could achieve more accurate quantitative prediction of soil total nitrogen content.

Neural Network: An Improved FCM for Multimodal Cultural Data Analysis

With the development of cloud computing, big data, and artificial intelligence (AI) technology, there is a growing interest in “cultural analysis.” Cultural analysis requires different types of data such as texts, pictures, and videos. The richness and differences of resources in the cultural field lead to diverse modalities of cultural data. Traditional text analysis methods can no longer meet the data analysis needs of current multimedia cultural resources. This article starts from cultural data’s feature information to solve the heterogeneity problem faced by massive multimodal cultural data analysis. It analyzes it from geography, time, art, and thematic character, classified and aggregated to form a multimodal cultural feature information matrix. The corresponding correlation measurement methods for different matrices from the above dimensions are proposed, solved in turn, and substituted into the optimized training back propagation (BP) neural network to obtain the final correlation degree. The improved fuzzy C-means (FCM) clustering algorithm is used to aggregate the high correlation cultural data based on the degree. The algorithm proposed in this study is compared with the existing algorithm. The experimental results show that the optimized BP neural network is at least 58% more accurate than the current method for calculating different matrices’ correlation degrees. In terms of accuracy, the improved fuzzy C-means algorithm effectively reduces the random interference in the selection of the initial clustering center, which is significantly higher than other clustering algorithms.