Traditional Back Propagation Neural Network Research Articles

A Ship Trajectory Prediction Method Based on an Optuna–BILSTM Model

In the field of maritime traffic management, overcoming the challenges of low prediction accuracy and computational inefficiency in ship trajectory prediction is crucial for collision avoidance. This paper presents an advanced solution using a deep bidirectional long- and short-term memory network (BILSTM) and the Optuna hyperparameter automatic optimized framework. Utilizing automatic identification system (AIS) data to analyze ship navigation patterns, the study applies Optuna to fine-tune the hyperparameters of the BILSTM network to improve prediction accuracy and efficiency. The developed Optuna–BILSTM model shows a remarkable 7% increase in prediction accuracy over traditional back propagation (BP) neural networks and standard BILSTM models. These results not only improve ship navigation and safety but also have significant implications for the development of autonomous ship collision avoidance systems, marking a significant step toward safer and more efficient maritime traffic management.

Analysis and Prediction of Urban Surface Transformation Based on Small Baseline Subset Interferometric Synthetic Aperture Radar and Sparrow Search Algorithm-Convolutional Neural Network-Long Short-Term Memory Model.

With the acceleration of urbanisation, urban areas are subject to the combined effects of the accumulation of various natural factors, such as changes in temperature leading to the thermal expansion or contraction of surface materials (rock, soil, etc.) and changes in precipitation and humidity leading to an increase in the self-weight of soil due to the infiltration of water along the cracks or pores in the ground. Therefore, the subsidence of urban areas has now become a serious geological disaster phenomenon. However, the use of traditional neural network prediction models has limitations when examining the causal relationships between time series surface deformation data and multiple influencing factors and when applying multiple influencing factors for predictive analyses. To this end, Sentinel-1A data from March 2017 to February 2023 were used as the data source in this paper, based on time series deformation data acquired using the small baseline subset interferometric synthetic aperture radar (SBAS-InSAR) technique. A sparrow search algorithm-convolutional neural network-long short-term memory (SSA-CNN-LSTM) neural network prediction model was built. The six factors of temperature, humidity, precipitation, and ground temperature at three different depths below the surface (5 cm, 10 cm, and 15 cm) were taken as the input of the model, and the surface deformation data were taken as the output of the neural network model. The correlation between the spatial and temporal evolution characteristics of the ground subsidence in urban areas and various influencing factors was analysed using grey correlation analysis, which proved that these six factors contribute to some extent to the deformation of the urban surface. The main urban area of Hohhot City, Inner Mongolia Autonomous Region, was used as the study area. In order to verify the efficacy of this neural network prediction model, the prediction effects of the multilayer perceptron (MLP), backpropagation (BP), and SSA-CNN-LSTM models were compared and analysed, with the values of the correlation coefficients of the feature points of A1, B1, and C1 being in the range of 0.92, 0.83, and 0.93, respectively. The results show that compared with the traditional MLP and BP neural network models, the SSA-CNN-LSTM model achieves a higher performance in predicting time series surface deformation data in urban areas, which provides new ideas and methods for this area of research.

Automated Grading of Angelica sinensis Using Computer Vision and Machine Learning Techniques

Angelica sinensis (Oliv.) Diels, a member of the Umbelliferae family, is commonly known as Danggui (Angelica sinensis, AS). AS has the functions of blood tonic, menstrual pain relief, and laxatives. Accurate classification of AS grades is crucial for efficient market management and consumer health. The commonly used method to classify AS grades depends on the evaluator’s observation and experience. However, this method has issues such as unquantifiable parameters and inconsistent identification results among different evaluators, resulting in a relatively chaotic classification of AS in the market. To address these issues, this study introduced a computer vision-based approach to intelligently grade AS. Images of AS at five grades were acquired, denoised, and segmented, followed by extraction of shape, color, and texture features. Thirteen feature parameters were selected based on difference and correlation analysis, including tail area, whole body area, head diameter, G average, B average, R variances, G variances, B variances, R skewness, G skewness, B skewness, S average, and V average, which exhibited significant differences and correlated with grades. These parameters were then used to train and test both the traditional back propagation neural network (BPNN) and the BPNN model improved with a growing optimizer (GOBPNN). Results showed that the GOBPNN model achieved significantly higher average testing precision, recall, F-score, and accuracy (97.1%, 95.9%, 96.5%, and 95.0%, respectively) compared to the BPNN model. The method combining machine vision technology with GOBPNN enabled efficient, objective, rapid, non-destructive, and cost effective AS grading.

Non-speech emotion recognition based on back propagation feed forward networks

Non-speech emotion recognition involves identifying emotions conveyed through non-verbal vocalizations such as laughter, crying, and other sound signals, which play a crucial role in emotional expression and transmission. This paper employs a nine-category discrete emotion model encompassing happy, sad, angry, peaceful, fearful, loving, hateful, brave, and neutral. A proprietary non-speech dataset comprising 2337 instances was utilized, with 384-dimensional feature vectors extracted. The traditional Backpropagation Neural Network (BPNN) algorithm achieved a recognition rate of 87.7% on the non-speech dataset. In contrast, the proposed Whale Optimization Algorithm - Backpropagation Neural Network (WOA-BPNN) algorithm, applied to a self-made non-speech dataset, demonstrated a remarkable accuracy of 98.6% . Notably, even without facial emotional cues, non-speech sounds effectively convey dynamic information, and the proposed algorithm excels in their recognition. The study underscores the importance of non-speech emotional signals in communication, especially with the continuous advancement of artificial intelligence technology. The abstract thus encapsulates the paper’s focus on leveraging AI algorithms for high-precision non-speech emotion recognition.

Accuracy Compensation Based on Optimized Neural Network for a Robotic Patient Positioning System with 6 Degrees of Freedom Used in Proton Therapy System

Proton therapy for tumor treatment is a typical application of nuclear technology. For proton therapy systems, robotic patient positioning systems (PPSs) are increasingly used because of their high flexibility and efficiency. Most robotic PPSs are developed based on industrial robots, which have good repeatability but low absolute position accuracy (1 to 3 mm) and do not satisfy the requirement of highly precise treatment. In this study, an optimized algorithm, named the Back Propagation Neural Network (BPNN) algorithm based on particle swarm optimization, is proposed to improve the performance of absolute positioning accuracy. A comparison of the training for the traditional BPNN and the optimized algorithm is presented. A series of experiments with different payload weights and tools is implemented to validate the performance of the proposed method. The training results show that the proposed method can improve the average predicted positioning error from 0.55 to 0.38 mm. The results of the experiment with a calibration tool show that the average position error is reduced from 4.10 to 0.32 mm. The results of the experiment with a carbon fiber couch top show that the average and maximal positioning errors are 0.35 and 0.77 mm, respectively. All the results verify the feasibility of the proposed method in this study in improving the position accuracy of the robotic PPS.

Groundwater chemical characteristic analysis and water source identification model study in Gubei coal mine, Northern Anhui Province, China

This study aims to accurately identify mine water sources and reduce the hazards caused by water inrush accidents in coal mines. Taking the Gubei coal mine as an example, the water quality results of the water samples from the Cenozoic unconsolidated aquifer, Permian sandstone fracture aquifer, and Carboniferous Taiyuan Formation limestone karst fracture aquifer in the mine area were tested, and K++Na+, Ca2+, Mg2+, Cl−, SO42−, HCO3−, TDS (Total Dissolved Solids), and pH were selected as the main indicators to study the water chemistry characteristics of the aquifer through water chemistry component analysis, major ion content analysis, Piper trilinear analysis, and correlation analysis. Thirty-five groups of water samples were randomly selected and imported into SPSS software for factor analysis (FA) and downsized to three main factors as the input variables of the artificial neural network model. The particle swarm optimization (PSO) code was written based on the MATLAB platform to improve the self-adjustment weights and acceleration factors for optimizing the initial weights and thresholds of the Back-Propagation (BP) neural network. The training and prediction samples were learned in the ratio of 8:2, and the recognition results were compared with the traditional BP neural network model. Results showed that the groundwater of the Gubei coal mine demonstrated a water quality vertical zoning pattern, and the chemical composition was dominated by cation K++Na+ and anion Cl−. The FA-PSO-BP neural network model has a higher accuracy of water source discrimination compared with the cluster analysis and the FA-BP neural network model. The FA-PSO-BP neural network model is worthy of further application in the problem of water source identification in mine water inrush.

Intrusion detection method based on improved social network search algorithm

The network security problem in today's world is becoming more and more prominent, and intrusion detection as a branch in the field of network security has been developed tremendously. At present, back propagation (BP) neural network is widely used in intrusion detection. However, its weights and thresholds are randomly initialized, so that fall into local optimal after training. To solve this problem, an intrusion detection model based on improved social network search (ISNS) algorithm to optimize BP neural network is proposed. The model first uses chaotic mapping and elite mechanism to improve the original Social Network Search (SNS) algorithm, and then uses the ISNS algorithm to filter the initial weights and thresholds of the BP neural network, and constructs the neural network with the optimal values to avoid falling into local optimum. The experimental results show that the proposed method solves the problem that the traditional BP neural network is easy to fall into local optimum. At the same time, the ISNS_BP model achieves better classification performance than other models, and the accuracy on the NSL-KDD and UNSW-NB15 datasets reaches 98.62% and 93.97%, respectively.

Prediction of Transformer Oil Temperature Based on an Improved PSO Neural Network Algorithm

Abstract: In addressing the issue of power transformer oil temperature prediction, traditional back propagation (BP) neural network algorithms have been found to suffer from local optimization and slow convergence. This study proposes an oil temperature prediction model based on an improved particle swarm optimization (PSO) neural network algorithm, which introduces an asymmetric adjustment learning factor and a mutation operator. The BP neural network, genetic algorithm (GA) optimization neural network, and the improved PSO neural network are compared by considering various factors, such as ambient temperature, load changes, and the number of cooler groups under different working conditions. Results show that the proposed algorithm improves the actual change trend of oil surface temperature and makes the transformer operation more stable to a certain extent. Background: The mathematical model for predicting transformer oil temperature is clear, but the parameters in the model are uncertain and vary with time. When subjected to different operating conditions, such as ambient temperature, load changes, and the number of cooler groups acting independently or in combination, the prediction results of the oil temperature model vary with different system parameters. Objective: This paper aims to enhance the accuracy of transformer temperature prediction. In order to optimize the oil temperature prediction model, asymmetric adjustment learning factors and mutant operators are added to meet diverse system parameter requirements. Methods: The paper utilizes an oil temperature prediction model based on an improved PSO neural network algorithm, which introduces an asymmetric adjustment learning factor and a mutation operator to address the limitations of the standard PSO algorithm. Results: This paper has employed a fusion algorithm of the genetic algorithm of the BP neural network and the PSO algorithm, and conducted simulation and experimental analysis. The simulation and experimental results demonstrate the accuracy and effectiveness of the fusion algorithm. Conclusion: This study demonstrates enhanced prediction accuracy of transformer oil surface temperature using the improved particle swarm optimization neural network algorithm. This algorithm has less prediction error under different working conditions compared to other algorithms. By increasing population diversity and combining inertia weights, the algorithm not only greatly improves its search performance but also avoids local optimization.

Evaluation of mine water quality based on the PCA–PSO–BP model

Abstract To enhance the mining area's overall use of mine water in the arid area of Western China and mitigate the current water scarcity problem, this paper introduces an intelligent optimization algorithm and neural network for mine water quality evaluation and proposes a principal component analysis (PCA)–particle swarm optimization (PSO)–back propagation (BP) mine water quality evaluation model. Firstly, the model uses PCA to identify the primary factors affecting mine water quality, then enhances the optimal weights and thresholds of the BP neural network based on the PSO algorithm, and the PCA–PSO–BP evaluation model with nine input layers, nine hidden layers, and one output layer is created. In addition, using the Shicaocun Mine as an example, the results demonstrate that the PCA–PSO–BP model has accurate mine water quality evaluation results, and the prediction accuracy reached 86.8255%. This exemplifies the PSO method's superiority to the BP neural network improvement. This study not only offers a novel theoretical framework for assessing and forecasting water quality in mining regions, but it also sets the stage for the possible broad use of state-of-the-art neural networks and optimization algorithms in the coal mining industry.

State of charge estimation for lithium‐ion battery pack based on real vehicle data and optimized backpropagation method by adaptive cross mutation sparrow search algorithm

AbstractIn response to the issues of traditional backpropagation (BP) neural networks in state of charge (SOC) estimation, including easy convergence to local optima, slow convergence speed, and low accuracy, this paper proposes a novel adaptive crossover mutation strategy and dynamic sparrow search algorithm to optimize BP networks' initial values and thresholds (ACMSSA‐BP). The proposed method is based on the sparrow search algorithm, where the number of producers and scroungers is adjusted through an adaptive factor. This improvement effectively transitions the search process from extensive full exploration to localized fine‐tuning search. In the position update phase of the producers, crossover mutation and dynamic search strategies are introduced to increase the diversity of good populations, prevent the algorithm from converging to local optima, and maintain its local search capability in the later stage. Using real transportation data from coal mining flame‐proof tracked vehicles, we applied correlation theory to extract model feature parameters and constructed a training data set to estimate the SOC. The results of both static and dynamic validation experiments have indicated that the ACMSSA‐BP method has delivered impressive performance in predicting SOC, as reflected in the mean absolute error, root mean squared error, and mean absolute percentage error values of less than 1.5%, 1.5%, and 1.6%, respectively. Compared with BP, SSA‐BP, CMSSA‐BP, PSO‐BP, and NARX_NN methods, the ACMSSA‐BP approach demonstrates enhanced accuracy in estimation, significant robustness, and impressive generalization capabilities.

Multi-source underwater DOA estimation using PSO-BP neural network based on high-order cumulant optimization

Due to the complex and changeable environment under water, the performance of traditional DOA estimation algorithms based on mathematical model, such as MUSIC, ESPRIT, etc., degrades greatly or even some mistakes can be made because of the mismatch between algorithm model and actual environment model. In addition, the neural network has the ability of generalization and mapping, it can consider the noise, transmission channel inconsistency and other factors of the objective environment. Therefore, this paper utilizes Back Propagation (BP) neural network as the basic framework of underwater DOA estimation. Furthermore, in order to improve the performance of DOA estimation of BP neural network, the following three improvements are proposed. (1) Aiming at the problem that the weight and threshold of traditional BP neural network converge slowly and easily fall into the local optimal value in the iterative process, PSO-BP-NN based on optimized particle swarm optimization (PSO) algorithm is proposed. (2) The Higher-order cumulant of the received signal is utilized to establish the training model. (3) A BP neural network training method for arbitrary number of sources is proposed. Finally, the effectiveness of the proposed algorithm is proved by comparing with the state-of-the-art algorithms and MUSIC algorithm.

Development of a Displacement Prediction System for Deep Excavation Using AI Technology

This manuscript delineates an innovative artificial intelligence-based methodology for forecasting the displacement of retaining walls due to extensive deep excavation processes. In our selection of 17 training cases, we strategically chose a wall configuration that was not influenced by the corner effects. This careful selection was conducted with the intention of ensuring that each deep excavation instance included in our study was supported symmetrically, thereby streamlining the analysis in the ensuing phases. Our proposed multilayer functional-link network demonstrates superior performance over the traditional backpropagation neural network (BPNN), excelling in the precise prediction of displacements at predetermined observation points, peak wall displacements, and their respective locations. Notably, the predictive accuracy of our advanced model surpassed that of the conventional BPNN and RIDO assessment tools by a substantial 5%. The network process model formulated through this research offers a valuable reference for future implementations in diverse geographical settings. Furthermore, by utilizing local datasets for the training, testing, and validation phases, our system ensures the effective and accurate execution of displacement predictions.

Determination of difluoromethane (R32) by non-dispersive infrared (NDIR) spectroscopy with temperature compensation by an improved whale optimization algorithm (IWOA)

To reduce the influence of temperature on difluoromethane (R32) gas sensors using non-dispersive infrared (NDIR), a temperature compensation method based on improved whale optimization algorithm (IWOA) and back propagation neural network (BPNN) is proposed. The IWOA that uses tent chaotic mapping to optimize the initial population of the whale optimization algorithm (WOA) and incorporates adaptive weights in the population iterations is used to optimize the weights and thresholds between nodes in each layer of the BPNN. The optimized BPNN is used to train the temperature compensation model. The experimental results show that the back propagation neural network optimized with improved whale optimization algorithm (IWOA-BPNN) reduces the maximum relative error to 1.33%, compared to 12.46 and 11.66% for the traditional BPNN and the back propagation neural network optimized with whale optimization algorithm (WOA-BPNN), respectively. IWOA-BPNN is also compared with previously reported algorithms, and the results show that the algorithm improves the accuracy of the sensor.

A friction temperature model for dynamic bearing operation based on neural network and genetic algorithm

This study used intelligent algorithms and experimental data to investigate the dynamic temperature of bearings. A test rig was created for angular contact ball bearings to measure the temperature of two bearings precisely while independently controlling rotational speed, axial load, radial load, and grease content. The study analyzed the steady temperature of bearings under different operating conditions using a friction temperature model based on SKF bearings. The bearings temperature was computed using traditional backpropagation neural network (BPNN) and genetic algorithms (GA). The study compared the results obtained through the SKF model and the actual temperature against those produced methods. GA displayed a higher level of accuracy and robustness. Upon evaluation, it was determined that both intelligent methods are suitable for processing variable data input. This article presents an intelligent approach to computing bearing temperature using statistical data analysis, which does not require analyzing dynamics, thermodynamics, or lubrication principles. The method's ability to compute transient temperature without boundary conditions or constraints makes it highly versatile. This feature renders it a valuable tool for bearing design and temperature control.

Fault diagnosis of a computer interlocking system for railway signal control

The signal control of the railway transportation system is crucial for operational safety. This paper briefly introduces the computer interlocking system for railway signal control, describes the tree-structured neural network used for fault diagnosis of the interlocking system, and introduces the particle swarm optimization (PSO) algorithm for improvement. Finally, a simulation experiment was conducted on a railway station to compare the traditional back-propagation neural network (BPNN), the support vector machine, the traditional tree-structured neural network, and the improved tree-structured neural network for fault diagnosis. It was found that the topological structure of the device distribution in the railway station could be transformed into a tree structure, and with the introduction of hidden nodes, it could become a binary tree structure where each leaf node represents a device; the improved tree-structured neural network had the highest recognition performance for both two-class tasks (determining system failure or not) and multi-class tasks (identifying fault type).

A Design of Apparel Appearance: Recognition and Evaluation of Clothing Pattern Styles under Deep Learning

Designing the appearance of clothing can effectively enhance its attractiveness and expand its marketability. This paper briefly introduces the Convolutional Neural Network (CNN) and applies it to the recognition and evaluation of clothing pattern styles to assist in evaluating clothing appearance design. A case analysis was then conducted. Firstly, the CNN algorithm was compared with the traditional Back-Propagation Neural Network (BPNN) algorithm, and then the design scheme proposed in this paper, called "Sanduo and Jiuru", was evaluated. The results showed that, compared to the BPNN algorithm, the CNN algorithm not only converged faster during training but also demonstrated superiority after the convergence became stable. In addition, the test set also verified the accuracy of the CNN algorithm in recognizing and evaluating clothing pattern styles. The evaluation of the "Sanduo and Jiuru" design was also very similar to human evaluation, and its excellence was analyzed accordingly.

Application of an intelligent evaluation model of online teaching based on improved BPNN

Online teaching quality evaluation is essential for university management, but the current evaluation model of online teaching quality has problems such as subjectivity, high randomness and slow convergence speed. Therefore, the study uses adaptive learning rate with momentum term for optimizing and improving the traditional back propagation neural network, and the improved adaptive back propagation neural network rating model has good application in small-scale low-dimensional datasets, and its computational power is weak in large-scale datasets. Therefore, the study uses deep noise reduction autoencoder to improve the model structure, and then obtains an improved backpropagation neural network. The model shows better performance in a self-made university online teaching evaluation dataset. When the quantity of hidden layers of the improved model is 2, the mean square error value of the model is 0.0026; when the quantity of hidden layer neurons is 25, the mean square error of the model is 0.0012; when the model is made comparison with the shallow model, the mean absolute percentage error, the mean square error, and the root mean square error perform the best with the values of 0.0502, 24.53, and 4.95, respectively. the experimental results verify that the experiment verified that the improved model possesses powerful computational ability, and also proved that it also possesses excellent forecasting convergence, which can better help universities to complete the evaluation of online teaching quality.

Effective fault module localization in substation critical equipment: an improved ant colony optimization and back propagation neural network approach

AbstractThe rapid development of substations has increased the demand for accurate and fast fault prediction systems. In order to achieve rapid localization and autonomous decision‐making of fault modules and types in substations, the article proposes a fault autonomous localization algorithm based on improved ant colony optimization (IACO) and back propagation neural network (BPNN). The fault data of the substation secondary equipment for training and testing the BPNN model is based on the actual operating equipment of the substation, which can significantly improve the reliability of the model results. In addition, the IACO is used to globally optimize the weights and thresholds of BPNN, and the number of hidden layer nodes in BPNN was analyzed to further improve the accuracy of the established fault prediction algorithm. The test results show that the fault prediction accuracy of the BPNN model optimized by IACO is 93.67%, which is significantly improved compared to the traditional BPNN and BPNN with ant colony optimization method (with an accuracy of 82.98% and 91.04%). The above results effectively demonstrate the high accuracy and effectiveness of the established prediction algorithm in processing data and locating faults, which can improve the maintenance and operational efficiency of substations.

Enhanced Neural Network for Rapid Identification of Crop Water and Nitrogen Content Using Multispectral Imaging

Precision irrigation and fertilization in agriculture are vital for sustainable crop production, relying on accurate determination of the crop’s nutritional status. However, there are challenges in optimizing traditional neural networks to achieve this accurately. This paper aims to propose a rapid identification method for crop water and nitrogen content using optimized neural networks. This method addresses the difficulty in optimizing the traditional backpropagation neural network (BPNN) structure. It uses 179 multi−spectral images of crops (such as maize) as samples for the neural network model. Particle swarm optimization (PSO) is applied to optimize the hidden layer nodes. Additionally, this paper proposes a double−hidden−layer network structure to improve the model’s prediction accuracy. The proposed double−hidden−layer PSO−BPNN model showed a 9.87% improvement in prediction accuracy compared with the traditional BPNN model. The correlation coefficient R2 for predicted crop nitrogen and water content was 0.9045 and 0.8734, respectively. The experimental results demonstrate high training efficiency and accuracy. This method lays a strong foundation for developing precision irrigation and fertilization plans for modern agriculture and holds promising prospects.

Combining the Back Propagation Neural Network and Particle Swarm Optimization Algorithm for Lithological Mapping in North China

Lithological mapping is a crucial tool for exploring minerals, reconstructing geological formations, and interpreting geological evolution. The study aimed to investigate the application of the back propagation neural network (BPNN) and particle swarm optimization (PSO) algorithm in lithological mapping. The study area is the Beiliutumiao map-sheet (No. K49E011021) in Inner Mongolia, China. This area was divided into two parts, with the left side used for training and the right side used for validation. Fifteen geological relevant factors, including geochemistry (1:200,000-scale) and geophysics (1:50,000-scale), were used as predictor variables. Taking one lithology as an example, the lithological binary mapping method was introduced in detail, and then the complete lithology was mapped. The model was compared with commonly used spatial data mining methods using the E-measure, S-measure, and Weighted F-measure values. In diorite testing, the accuracy and kappa of the optimized model were 92.11% and 0.81, respectively. The validation results showed that our method outperformed the traditional BPNN and weights-of-evidence approaches. In the extension of the complete lithological mapping, the accuracy, recall, and F1-score were 82.66%, 74.54%, and 0.76, respectively. Thus, the proposed method is useful for predicting the distribution of one lithology and completing the whole lithological mapping at a fine scale. In addition, the trained network can be extended to an adjacent area with similar lithological features.