Neural Network Prediction Method Research Articles

Construction of data base containing flow field information under different blade placement of pumping station in digital twin

Abstract The data base in digital twin technology includes various data such as the structure, performance, and operating status of the physical object. And the data base is the key point for the construction of the digital twin system. However, the data bases of different digital twin systems usually contain different information. In this paper, the flow state of the pumping station under different angles of four blades is selected as the research object. The research methods in this paper mainly include Latin hypercube sampling, numerical simulation and neural network prediction. The Latin hypercube sampling is used to select the typical Angle difference model. The numerical simulation is used to obtain flow field information. The neural network prediction method is used to construct a complete data base. The study incorporates an actual large-scale mixed-flow pump station to head, efficiency, and shaft power as the prediction targets. It utilizes the different angle difference data of ten sets of four blades as the training set, forty sets of remaining samples as the experiment set, and employs the backpropagation neural network method for constructing the prediction network. By changing the experiment set and test set, the method can efficiently complete the construction of data base including actual test and numerical simulation, which provides a solid foundation for the construction of digital twin platform.

A deep neural network prediction method for diabetes based on Kendall's correlation coefficient and attention mechanism.

Diabetes is a chronic disease, which is characterized by abnormally high blood sugar levels. It may affect various organs and tissues, and even lead to life-threatening complications. Accurate prediction of diabetes can significantly reduce its incidence. However, the current prediction methods struggle to accurately capture the essential characteristics of nonlinear data, and the black-box nature of these methods hampers its clinical application. To address these challenges, we propose KCCAM_DNN, a diabetes prediction method that integrates Kendall's correlation coefficient and an attention mechanism within a deep neural network. In the KCCAM_DNN, Kendall's correlation coefficient is initially employed for feature selection, which effectively filters out key features influencing diabetes prediction. For missing values in the data, polynomial regression is utilized for imputation, ensuring data completeness. Subsequently, we construct a deep neural network (KCCAM_DNN) based on the self-attention mechanism, which assigns greater weight to crucial features affecting diabetes and enhances the model's predictive performance. Finally, we employ the SHAP model to analyze the impact of each feature on diabetes prediction, augmenting the model's interpretability. Experimental results show that KCCAM_DNN exhibits superior performance on both PIMA Indian and LMCH diabetes datasets, achieving test accuracies of 99.090% and 99.333%, respectively, approximately 2% higher than the best existing method. These results suggest that KCCAM_DNN is proficient in diabetes prediction, providing a foundation for informed decision-making in the diagnosis and prevention of diabetes.

Research on indoor and outdoor positioning switching algorithm based on improved PSO-BP

Aiming at the problems of poor positioning accuracy in the indoor and outdoor junction areas and the loss of positioning signals and discontinuous positioning results during the transition from the indoor area to the outdoor area, this paper proposes a machine learning method to solve the positioning problem in the indoor and outdoor junction areas and the switching problem of positioning methods. An indoor and outdoor positioning switching algorithm based on Particle Swarm Optimization-Back Propagation (PSO-BP) and BP neural network is designed. Through this algorithm, the position of the positioning tag can be judged independently and the coordinates of the positioning tag in the indoor and outdoor junction area can be predicted independently. The experimental results show that the accuracy of positioning area judgment based on PSO-BP neural network can reach 99.91%. The indoor and outdoor boundary area coordinates obtained by BP neural network prediction method are lower than the root mean square error of ultra wide band indoor positioning method and BDS positioning method. The algorithm model proposed in this paper effectively improves the positioning problem of the indoor and outdoor junction area and the positioning problem of the transition from the indoor area to the outdoor area, improves the positioning accuracy and positioning stability of the indoor and outdoor junction area, reduces the positioning cost, and has strong practicability.

Tension identification of bridge suspension rods based on multiple impedance parameters from self-inductive coils

ABSTRACT This study introduces a method for identifying tension in suspension rods based on impedance from self-inductive coils, grounded in electromagnetic induction and magnetic elasticity principles. The theoretical feasibility of this method was analysed through a high-frequency inductive coil model. Steel strand tensile tests, utilising self-inductive coil sensors at various frequencies, explored the relationship between resistance, inductance, and capacitance with tension. According to the experimental results, the sensitivity coefficient is used to determine the optimal prediction frequency of each impedance parameter. Within the optimal frequency range, a single impedance parameter regression prediction method based on the impedance difference ratio index was proposed. To improve the accuracy and stability of the predictions, a GA-BP neural network prediction model based on the fusion of multiple impedance parameters was proposed. The results indicate that the GA-BP neural network prediction method, which integrates multiple impedance parameters, achieves higher accuracy in identifying cable forces than the single impedance parameter regression prediction method. Specifically, at a frequency of 5 kHz, the relative error of the GA-BP neural network prediction method, which integrates multiple impedance parameters, was only 3.74%. This research offers a novel approach for future in-service tension identification in bridge suspension rods.

Prediction of dredged soil settlement based on improved BP neural network

Based on the measured settlement data from the Reclamation Project of Eastern Hengsha Shoal, two improved BP neural networks optimized by the particle swarm optimization (PSO) algorithm and genetic algorithm (GA) respectively are employed to predict the long-term consolidation settlement of the dredged soil in the reclamation area. By comparing the prediction results of the improved neural networks with those obtained from standard BP neural network and various curve fitting methods including hyperbola method, Hoshino method, and Asaoka method, as well as field-measured data, the improved neural network methods are validated to have a better predictive performance. The error analysis results indicate that compared with the standard BP neural network and curve-fitting prediction methods, the improved BP neural networks exhibit smaller prediction errors in terms of MAE, MSE, RMSE, MAPE, etc., and have a closer match between the predicted settlement values and the actual values. The selected GA-BP and PSO-BP neural network prediction methods in this study demonstrate higher prediction accuracy. Meanwhile, GA-BP demonstrates higher prediction accuracy and generalization ability than PSO-BP, but PSO-BP shows faster convergence in terms of prediction speed. The preferred methods presented in this paper exhibit good robustness and high reliability, which can achieve accurate prediction of dredged soil settlement. Additionally, unlike the traditional curve-fitting methods, the selected methods can effectively reduce the influence of subjective factors during the settlement prediction, and provide a new approach for settlement prediction, which can be effectively utilized as a substitute for long-term settlement monitoring and give reference for the design of settlement control of foundation of further construction projects.

Shear Wave Velocity Prediction Based on the Long Short-Term Memory Network with Attention Mechanism

Shear wave velocity (VS) is a vital prerequisite for rock geophysics. However, due to historical, cost, and technical reasons, the shear wave velocity of some wells is missing. To reduce the deviation of the description of underground oil and gas distribution, it is urgent to develop a high-precision neural network prediction method. In this paper, an attention module is designed to automatically calculate the weight of each part of the input value. Then, the weighted data are fed into the long short-term memory network to predict shear wave velocities. Numerical simulations demonstrate the efficacy of the proposed method, which achieves a significantly lower MAE of 38.89 compared to the LSTM network’s 45.35 in Well B. In addition, the relationship between network input length and prediction accuracy is further analyzed.

Corrigendum: Research on neural network prediction method for upgrading scale of natural gas reserves

Corrigendum: Research on neural network prediction method for upgrading scale of natural gas reserves

Deep learning model optimization of 110 kV line ice-melting technology without power failure

Abstract In order to achieve ice disaster warning for transmission lines, this article proposes an improved method based on particle swarm optimization (PSO) backpropagation (BP) neural network for ice thickness prediction. The study selects multiple factors such as temperature, humidity, and wind speed, introducing neural network prediction methods to explore the situation of ice thickness on transmission lines. The predictions from this model demonstrate closer alignment with actual values and achieve superior prediction accuracy compared to both the BP model and the PSO-BP model, and the average absolute percentage error of improved PSO (IPSO)-BP is 0.007. Using a 110 kV line as the experimental object, the study conducts ice melting without power outage by changing of the grid currents. Additionally, this predictive model method is employed for ice warnings, assessing ice coverage levels by computing the ice coverage ratio. This facilitates precise control over the activation and deactivation of ice melting devices. The method proposed in this study addresses the issue of low accuracy resulting from the singular data types used in traditional early warning models. Future efforts should focus on further validating the applicability of this method under varying climatic and environmental conditions to achieve real-time, precise control over line ice melting.

Supply Side Reform of Tourism in Ethnic Areas Empowers Common Wealth--Based on the Perspective of Tourism Human Resources

Abstract As part of the supply-side reform, tourism in ethnic areas is increasingly recognized as crucial for achieving shared prosperity. This paper introduces a model from the perspective of tourism human resources designed to predict the demand for such resources. The model aims to inform strategies that support supply-side reforms in tourism for ethnic areas. The GM(1,1) model is constructed with the gray algorithm model, and on the basis of the gray algorithm model, the gray prediction method is effectively combined with the BP neural network prediction method, and the prediction preferences are made according to the variance as well as the prediction method idea of the preferred combination, and the final prediction results are obtained. In the analysis of the tourism human resources profile and demand forecast in Xinjiang, the total number of tourism employment in 2022 accounts for 17.22% of the total number of jobs, and the total number of tourism students in school reaches 6,189. Only 11.66% of tourism human resources are highly educated talents with a bachelor’s degree or above, and the number of personnel with senior titles is only 8.42%, which is obviously low in high-education and high-level talents. Nearly 80% of the personnel in non-tourism-related categories are still there, and the specialization of tourism management personnel is low. The demand for tourism human resources in Xinjiang in 2027 is expected to reach 2.1179 million.

Application of Artificial Neural Network in Predicting Direct Economic Losses Due to Earthquake

Accurately predicting the direct economic losses caused by earthquakes is important for policy makers for disaster budgets. Before a disaster strikes, it is important to consider the public policy costsassociated with disaster relief and recovery. The aim of this study is to provide a risk assessment approach, which can benefit all parties involved. Artificial neural networks are widely used for time seriesforecasting, especially financial forecasting. Therefore, this study proposes a cutting-edge forecastingmethod such as backpropagation neural network (BPNN) and other prediction methods: neural networkautoregressive (NNAR) and ARIMA-GARCH to obtain the best prediction results. This paper appliesinterpolation data to increase the amount of data used. Two interpolations were applied to amplify theoriginal small sample with virtual points, namely cubic splines and further piecewise interpolation using. The results of this study are the cubic spline interpolation is the most effective way to solve thesmall sampling problem to predict direct economic losses due to the Indonesian earthquake and theBPNN method outperforms other traditional methods with an RMSE of 0.024 in the training period and0.174 in the testing period, significantly lower than other methods. The results of this research can beused as reference material for the government in estimating the level of earthquake losses and can beused to develop risk reduction strategies

Wind Power Analysis Using Machine Learning in Wind Turbines

Wind Power Analysis Using Machine Learning in Wind Turbines

Research on neural network prediction method for upgrading scale of natural gas reserves

With the gradual decline of natural gas production, reserve upgrading has become one of the important issues in natural gas exploration and development. However, the traditional reserve upgrade forecasting method is often based on experience and rules, which is subjective and unreliable. Therefore, a prediction method based on neural network is proposed in this paper to improve the accuracy and reliability of reserve upgrade prediction. In order to achieve this goal, by collecting the relevant data of natural gas exploration and development in Sichuan Basin, including geological parameters, production parameters and other indicators, and processing and analyzing the data, the relevant characteristics of reserves increase are extracted. Then, a neural network model based on multi-layer perceptron (MLP) is constructed and trained and optimized using backpropagation algorithm. The results show that the prediction accuracy of the constructed neural network model can reach more than 90% and can effectively predict the reserve upgrading. Experiments show that the model has high accuracy and reliability, and is significantly better than the traditional prediction methods. The method has good stability and reliability, and is suitable for a wider range of natural gas fields.

Multi-attribute Probabilistic Neural Network Reservoir Prediction Based on Mudstone Acoustic Time Difference Decompaction Correction

In order to accurately predict reservoirs with similar physical properties between sandstone and surrounding rocks, this study takes Carboniferous-Permian system in southeastern Ordos Basin as an example, and puts forward a multi-attribute probabilistic neural network prediction method based on acoustic time difference compaction correction. Firstly, the time-frequency analysis method is used to divide the frequency of the original acoustic time difference curve, and the low-frequency long-trend correction of all wells is carried out based on the low-frequency components of standard wells. Then, it is frequency-divided and fused with the original high-frequency components to form new acoustic time difference data, and the long-term trend decompaction correction of acoustic time difference is completed, so that the overlapping area of acoustic time difference between reservoir and surrounding rock is reduced. Multi-attribute probabilistic neural network is an algorithm based on sampling points, before reservoir prediction, inversion parameters must be optimized to save calculation time and reduce prediction error. There is a high correlation between longitudinal velocity and lithologic change trend on the inversion profile of sound velocity. The comparison results of various inversions show that the velocity profile of multi-attribute probabilistic neural network has the highest coincidence rate and high resolution with the velocity curve of logging acoustic wave. The example shows that this method has high accuracy in predicting reservoir thickness, especially for the strata with little difference in physical properties between reservoir and surrounding rock.

Short-Term Wind Power Prediction Based on CEEMDAN-SE and Bidirectional LSTM Neural Network with Markov Chain

Accurate wind power data prediction is crucial to increase wind energy usage since wind power data are characterized by uncertainty and randomness, which present significant obstacles to the scheduling of power grids. This paper proposes a hybrid model for wind power prediction based on complementary ensemble empirical mode decomposition with adaptive noise (CEEMDAN), sample entropy (SE), bidirectional long short-term memory network (BiLSTM), and Markov chain (MC). First, CEEMDAN is used to decompose the wind power series into a series of subsequences at various frequencies, and then SE is employed to reconstruct the wind power series subsequences to reduce the model’s complexity. Second, the long short-term memory (LSTM) network is optimized, the BiLSTM neural network prediction method is used to predict each reconstruction component, and the results of the different component predictions are superimposed to acquire the total prediction results. Finally, MC is used to correct the model’s total prediction results to increase the accuracy of the predictions. Experimental validation with measured data from wind farms in a region of Xinjiang, and computational results demonstrate that the proposed model can better fit wind power data than other prediction models and has greater prediction accuracy and generalizability for enhancing wind power prediction performance.

A Double-Layer Vehicle Speed Prediction Based on BPNN-LSTM for Off-Road Vehicles.

The accurate prediction of vehicle speed is crucial for the energy management of vehicles. The existing vehicle speed prediction (VSP) methods mainly focus on road vehicles and rarely on off-road vehicles. In this paper, a double-layer VSP method based on backpropagation neural network (BPNN) and long short-term memory (LSTM) for off-road vehicles is proposed. First of all, considering the motion characteristics of off-road vehicles, the VSP problem is established and the relationship between the variables in the problem is carefully analyzed. Then, the double-layer VSP framework is presented, which consists of speed prediction and information update layers. The speed prediction layer established by using LSTM is to predict vehicle speed in the horizon, and the information update layer built by BPNN is to update the prediction information. Finally, with the help of mining truck and loader operation scenarios, the proposed VSP method is compared with the analytical method, BPNN prediction method, and recurrent neural network (RNN) prediction method in terms of speed prediction accuracy. The results show that, under the premise of ensuring the real-time prediction performance, the average prediction error of the proposed BPNN-LSTM prediction method under two operation scenarios reduces by 48.14%, 35.82% and 30.09% compared with the other three methods, respectively. The proposed speed prediction method provides a new solution for predicting the speed of off-road vehicles, effectively improving the speed prediction accuracy.

On modelling Nd:Yag nanosecond laser milling process by neural network and multi response prediction methods

In this paper, the milling effects on AISI 316 L steel by nanosecond pulsed laser was studied under different parameters as laser pulse peak power, laser repetition rate, scanning speed and focal point. Based on a full factorial design, 108 tests were conducted. MD varied from 15.58 to 60.14 µm; MRR from 0.0028 to 0.0195 mm3/s; Ra from 0.13 to 1.49 µm. The study showed that the laser power and scanning speed have a great influence on material removal. When the laser spot was focused on the upper sample surface, the removing material process was improved. Moreover, Response Surface Methodology and Artificial Neural Network were used to analyze and predict the milling depth, the material removal rate, and the surface roughness. The predictive capabilities of the methods were compared. The predicted R-square for RMS model were 99.70%, 77.79% and 67.73%, while for FF-BPNN were 99.99%, 99.85% and 81.98% for MD, MRR and Ra respectively. Therefore, both models demonstrated strong ability to match the responses. Neural network model exhibited superior predictive power.

Prediction of tool wear in milling process based on BP neural network optimized by firefly algorithm

Long-time discontinuous contact is easy to cause tool wear during milling. To decrease the impact of severe wear on workpiece quality and processing efficiency, cutting tools should be replaced timely. Therefore, tool wear prediction is an important aspect in improving process efficiency, ensuring machining precision and realizing intelligent manufacturing. To boost the precision of online prediction of tool wear, this paper suggests a novel approach to monitor tool wear by optimizing backpropagation (BP) neural network via firefly algorithm (FA). Specifically, the progressive semi-soft threshold function is applied to the process of cutting force signal noise reduction, which reduces redundant signals and noise interference in the signal. Time-domain analysis, frequency-domain analysis, and wavelet packet decomposition are utilized, cutting force features are extracted, and Pearson correlation coefficient is used to sift out signal features that are highly connected to tool wear. The FA is used to improve the BP neural network's weights and thresholds. Through learning nonlinear mapping, relationship between tool flank wear and signal features is realized. A prediction model of tool wear of FA-BP is constructed. Milling experiments validate the prediction model in the milling process. The experimental outcomes confirm the precision and reliability of the method. In comparison to BP neural network, genetic algorithm optimized BP neural network and particle swarm optimization algorithm optimized BP neural network prediction method, it has a greater prediction accuracy and a stronger training impact, and has superior performance. The research results can give a theoretical foundation and technological assistance for predicting tool wear, which is crucial for improving workpiece quality, processing efficiency, and promoting intelligent development.

Photovoltaic Power-Stealing Identification Method Based on Similar-Day Clustering and QRLSTM Interval Prediction

In order to defraud state subsidies, some unscrupulous users use improper means to steal photovoltaic (PV) power. This behavior brings potential safety hazards to photovoltaic grid-connected operations. In this paper, a photovoltaic power-stealing identification method based on similar-day clustering and interval prediction of the quantile regression model for long short-term memory neural network (QRLSTM) is proposed. First, photovoltaic data are clustered into three similar days by the similar-day clustering according to weather conditions. Second, compared with the quantile regression neural network (QRNN) prediction method, the good prediction performance of the QRLSTM method is illustrated. Third, using the prediction intervals with different confidence levels on three similar days, according to the time scale (short-term, medium-term and long-term) combined with different electricity-stealing judgment indicators, a three-layer photovoltaic power-stealing screening framework is constructed, and the degree of user power stealing is qualitatively analyzed. Last, the power generation data of eight photovoltaic users in a certain region of northwest China and the data of four groups of artificially constructed power-stealing users are used as an example for simulation. The simulation results prove the feasibility of the proposed method in this paper.

Causes and Impacts of Decreasing Chlorophyll-a in Tibet Plateau Lakes during 1986–2021 Based on Landsat Image Inversion

Lake chlorophyll-a (Chl-a) is one of the important components of the lake ecosystem. Numerous studies have analyzed Chl-a in ocean and inland water ecosystems under pressures from climate change and anthropogenic activities. However, little research has been conducted on lake Chl-a variations in the Tibet Plateau (TP) because of its harsh environment and limited opportunities for in situ data monitoring. Here, we combined 95 in situ measured lake Chl-a concentration data points and the Landsat reflection spectrum to establish an inversion model of Chl-a concentration. For this, we retrieved the mean annual Chl-a concentration in the past 35 years (1986–2021) of 318 lakes with an area of > 10 km2 in the TP using the backpropagation (BP) neural network prediction method. Meteorological and hydrological data, measured water quality parameters, and glacier change in the lake basin, along with geographic information system (GIS) technology and spatial statistical analysis, were used to elucidate the driving factors of the Chl-a concentration changes in the TP lakes. The results showed that the mean annual Chl-a in the 318 lakes displayed an overall decrease during 1986–2021 (−0.03 μg/L/y), but 63%, 32%, and 5% of the total number exhibited no significant change, significant decrease, and significant increase, respectively. After a slight increase during 1986–1995 (0.05 μg/L/y), the mean annual lake Chl-a significantly decreased during 1996–2004 (−0.18 μg/L/y). Further, it decreased slightly during 2005–2021 (−0.02 μg/L/y). The mean annual lake Chl-a concentration was significantly negatively correlated with precipitation (R2 = 0.48, p < 0.01), air temperature (R2 = 0.31, p < 0.01), lake surface water temperature (LSWT) (R2 = 0.51, p < 0.01), lake area (R2 = 0.42, p < 0.01), and lake water volume change (R2 = 0.77, p < 0.01). The Chl-a concentration of non-glacial-meltwater-fed lakes were higher than those of glacial-meltwater-fed lakes, except during higher precipitation periods. Our results shed light on the impacts of climate change on Chl-a variation in the TP lakes and lay the foundation for understanding the changes in the TP lake ecosystem.

Deep Learning Framework of Convolutional Neural Network (CNN) and Attention CNN for Early Diagnosis of Alzheimer's Disease

One of the biggest killers in the industrialized world is Alzheimer's disease (AD). Although computer-aided techniques have shown promising outcomes in laboratory experiments, they have yet to be used in a clinical setting. Recently, deep neural networks have gained traction, particularly for image processing tasks. There has been a dramatic increase in the number of publications written on the topic of identifying AD using deep learning since 2017. It has been observed that deep networks are more efficient than standard machine learning methods for detecting AD. It remains difficult to identify AD because distinguishing between comparable brain signals during categorization needs an extremely discriminative depiction of features. This paper proposed a deep neural network method for prediction the AD. Low-level computer vision has been a hotspot for research into deep convolutional neural networks (CNNs). Studies often focus on enhancing performance through the use of very deep CNNs. Yet, as one goes deeper, the effect of the shallow layers on the deeper ones gradually diminishes. Prompted by reality. This paper compared with the CNN and attention CNN models. The proposed model applied in the AD dataset which contains 5121 images for the train set. The results showed the attention CNN model is better than the CNN model in accuracy, precision, recall, loss, and AUC.