Research on the Prediction Method of Centrifugal Pump Performance Based on a Double Hidden Layer BP Neural Network

Development and validation of machine learning models for MASLD: based on multiple potential screening indicators

Background

Multifaceted factors play a crucial role in the prevention and treatment of metabolic dysfunction-associated steatotic liver disease (MASLD). This study aimed to utilize multifaceted indicators to construct MASLD risk prediction machine learning models and explore the core factors within these models.

Methods

MASLD risk prediction models were constructed based on seven machine learning algorithms using all variables, insulin-related variables, demographic characteristics variables, and other indicators, respectively. Subsequently, the partial dependence plot(PDP) method and SHapley Additive exPlanations (SHAP) were utilized to explain the roles of important variables in the model to filter out the optimal indicators for constructing the MASLD risk model.

Results

Ranking the feature importance of the Random Forest (RF) model and eXtreme Gradient Boosting (XGBoost) model constructed using all variables found that both homeostasis model assessment of insulin resistance (HOMA-IR) and triglyceride glucose-waist circumference (TyG-WC) were the first and second most important variables. The MASLD risk prediction model constructed using the variables with top 10 importance was superior to the previous model. The PDP and SHAP methods were further utilized to screen the best indicators (including HOMA-IR, TyG-WC, age, aspartate aminotransferase (AST), and ethnicity) for constructing the model, and the mean area under the curve value of the models was 0.960.

Conclusions

HOMA-IR and TyG-WC are core factors in predicting MASLD risk. Ultimately, our study constructed the optimal MASLD risk prediction model using HOMA-IR, TyG-WC, age, AST, and ethnicity.

Estimating the methane flux of the Dajiuhu subalpine peatland using machine learning algorithms and the maximal information coefficient technique

The eddy covariance (EC) technique has emerged as the most widely used method for long-term continuous methane flux (FCH4) observations. However, the completeness of the FCH4 time series is limited by instrumental failures and data quality issues, resulting in missing data gaps ranging from 20 % to 90 %. In this situation, the excellent performance of machine learning (ML) algorithms in filling missing FCH4 data has provided a foundation for developing regional-scale FCH4 models. In this study, we established estimation models for FCH4 utilizing random forest (RF), support vector machine (SVM), back propagation (BP) and nonlinear multiple regression (MLR) algorithms. The maximal information coefficient (MIC) technique was employed to identify and rank the environmental factors that were correlated with FCH4. Our findings revealed that soil temperature (Ts), soil water content (SWC) and air temperature (Ta) were the primary environmental factors influencing FCH4. Among the four algorithms, from perspectives of model accuracy and relatively small number of driving factors, the RF models exhibited the best performance, followed by BP and SVM, whereas MLR demonstrated the lowest performance. Among the 144 RF models established using nine datasets, RF model with 8 driving factors in all-year (RFall-year8) could capture seasonal variations. Ultimately, we recommend (RFall-year8 as the optimal model for estimating FCH4 in the Dajiuhu subalpine peatland.

Research on the path of building carbon peak in China based on LMDI decomposition and GA-BP model

The building sector contributes significantly to carbon emissions, impeding China's progress toward its 2030 carbon emissions peak target due to the limited utilization of renewable energy sources. This study aims to forecast the peak and timing of carbon emissions in China's construction industry to chart a low-carbon roadmap for the sector's future. Initially, an extended logarithmic mean divisia index (LMDI) decomposition model, based on the Kaya identity, is proposed to gauge the contribution levels of driving factors affecting building carbon intensity. Subsequently, a hybrid prediction model (IGA-BP) is constructed, employing an optimized two-hidden-layer neural network via a genetic algorithm, to forecast building carbon emissions and intensity. Additionally, four scenarios are outlined, each defining pathways to simulate emissions peak, carbon peak timing, and intensity within the Chinese building sector from 2020 to 2050. The research findings reveal: (1) The final emission factor of buildings primarily drives the surge in building carbon intensity, while the industrial structure stands as the most significant limiting factor. (2) Compared to alternative models, the proposed hybrid prediction model more effectively captures the evolution pattern of carbon emissions. (3) The prediction results indicate that China's building carbon intensity has reached its peak. Pathway 12 closely aligns with the sector's carbon emissions peak, projecting a peak value of 5.609 billion tons in 2029. To attain this pathway, China needs to develop more precise and feasible emission reduction strategies for its buildings. Overall, the research outcomes furnish robust references for decision-making in future efforts aimed at reducing building emissions.

Corrosion Degree Evaluation of Polymer Anti-Corrosive Oil Well Cement under an Acidic Geological Environment Using an Artificial Neural Network

Oil well cement is prone to corrosion and damage in carbon dioxide (CO2) acidic gas wells. In order to improve the anti-corrosion ability of oil well cement, polymer resin was used as the anti-corrosion material. The effect of polymer resin on the mechanical and corrosion properties of oil well cement was studied. The corrosion law of polymer anti-corrosion cement in an acidic gas environment was studied. The long-term corrosion degree of polymer anti-corrosion cement was evaluated using an improved neural network model. The cluster particle algorithm (PSO) was used to improve the accuracy of the neural network model. The results indicate that in acidic gas environments, the compressive strength of polymer anti-corrosion cement was reduced under the effect of CO2, and the corrosion depth was increased. The R2 of the prediction model PSO-BPNN3 is 0.9970, and the test error is 0.0136. When corroded for 365 days at 50 °C and 25 MPa pressure of CO2, the corrosion degree of the polymer anti-corrosion cement was 43.6%. The corrosion depth of uncorroded cement stone is 76.69%, which is relatively reduced by 33.09%. The corrosion resistance of cement can be effectively improved by using polymer resin. Using the PSO-BP neural network to evaluate the long-term corrosion changes of polymer anti-corrosion cement under complex acidic gas conditions guides the evaluation of its corrosion resistance.

A Deep Learning Approach for Predicting Multiple Sclerosis

This paper proposes a deep learning model based on an artificial neural network with a single hidden layer for predicting the diagnosis of multiple sclerosis. The hidden layer includes a regularization term that prevents overfitting and reduces the model complexity. The purposed learning model achieved higher prediction accuracy and lower loss than four conventional machine learning techniques. A dimensionality reduction method was used to select the most relevant features from 74 gene expression profiles for training the learning models. The analysis of variance test was performed to identify the statistical difference between the mean of the proposed model and the compared classifiers. The experimental results show the effectiveness of the proposed artificial neural network.

Predicting Centrifugal Pumps’ Complete Characteristics Using Machine Learning

The complete characteristics of centrifugal pumps are crucial for the modeling of hydraulic transient phenomena occurring in pipe systems. However, due to the effort required to obtain these curves, pump manufacturers typically only provide basic information, particularly when the pump operates under normal conditions. To acquire the full characteristic curves based on the manufacturer’s normal performance curve, a machine learning (ML) model is proposed to predict full, complete Suter curves using a pump’s specific speed with the known parts of the Suter curve. The training data for the model are sourced from the available Suter curves from laboratory experiments. Subsequently, the proposed ML model combines several types of regression models in an attempt to find the most accurate prediction in terms of the root mean square error (RMSE). The result proved highly efficient, as the experiments attained a maximum RMSE value of 0.032 across the three categories of centrifugal pumps based on their specific speeds, hence demonstrating the potential of machine learning in the study of pump characteristic curves.

The Estimation of Centrifugal Pump Flow Rate Based on the Power–Speed Curve Interpolation Method

During the global energy crisis, it is essential to improve the energy efficiency of pumps by adjusting the pump’s control strategy according to the operational states. However, monitoring the pump’s operational states with the help of external sensors brings both additional costs and risks of failure. This study proposed an interpolation method based on PN curves (power–speed curves) containing information regarding motor shaft power, speed, and flow rate to achieve high accuracy in predicting the pump’s flow rates without flow sensors. The impact factors on the accuracy of the estimation method were analyzed. Measurements were performed to validate the feasibility and robustness of the PN curve interpolation method and compared with the QP and back-propagation neural network (BPNN) methods. The results indicated that the PN curve interpolation method has lower errors than the other two prediction models. Moreover, the average absolute errors of the PN curve interpolation method in the project applications at 47.5 Hz, 42.5 Hz, 37.5 Hz, and 32.5 Hz are 0.1442 m3/h, 0.2047 m3/h, 0.2197 m3/h, and 0.1979 m3/h. Additionally, the average relative errors are 2.0816%, 3.2875%, 3.6981%, and 2.9419%. Hence, this method fully meets the needs of centrifugal pump monitoring and control.

Two-Stage Hybrid Model for Efficiency Prediction of Centrifugal Pump

Accurately predict the efficiency of centrifugal pumps at different rotational speeds is important but still intractable in practice. To enhance the prediction performance, this work proposes a hybrid modeling method by combining both the process data and knowledge of centrifugal pumps. First, according to the process knowledge of centrifugal pumps, the efficiency curve is divided into two stages. Then, the affinity law of pumps and a Gaussian process regression (GPR) model are explored and utilized to predict the efficiency at their suitable flow stages, respectively. Furthermore, a probability index is established through the prediction variance of a GPR model and Bayesian inference to select a suitable training set to improve the prediction accuracy. Experimental results show the superiority of the hybrid modeling method, compared with only using mechanism or data-driven models.

Energy Performance Curves Prediction of Centrifugal Pumps Based on Constrained PSO-SVR Model

It is of great significance to predict the energy performance of centrifugal pumps for the improvement of the pump design. However, the complex internal flow always affects the performance prediction of centrifugal pumps, particularly under low-flow operating conditions. Relying on the data-fitting method, a multi-condition performance prediction method for centrifugal pumps is proposed, where the performance relationship is incorporated into the particle swarm optimization algorithm, and the prediction model is optimized by automatically meeting the performance constraints. Compared with the experimental results, the performance under multiple operating conditions is well predicted by introducing performance constraints with the mean absolute relative error (MARE) for the head, power and efficiency of 0.85%, 1.53%,1.15%, respectively. By comparing the extreme gradient boosting and support vector regression models, the support vector regression is more suitable for the prediction of performance curves. Finally, by introducing performance constraints, the proposed model demonstrates a dramatic decrease in the head, power and efficiency of MARE by 98.64%, 82.06%, and 85.33%, respectively, when compared with the BP neural network.

Prediction of the Deformation of Aluminum Alloy Drill Pipes in Thermal Assembly Based on a BP Neural Network

The connection between the steel joint and aluminum alloy pipe is the weak part of the aluminum alloy drill pipe. Practically, the interference connection between the aluminum alloy rod and the steel joint is usually realized by thermal assembly. In this paper, the relationship between the cooling water flow rate, initial heating temperature and the thermal deformation of the steel joint in interference thermal assembly was studied and predicted. Firstly, the temperature data of each measuring point of the steel joint were obtained by a thermal assembly experiment. Based on the theory of thermoelasticity, the analytical solution of the thermal deformation of the steel joint was studied. The temperature function was fitted by the least square method, and the calculated value of radial thermal deformation of the section was finally obtained. Based on the BP neural network algorithm, the thermal deformation of steel joint section was predicted. Besides, a prediction model was established, which was about the relationship between cooling water flow rate, initial heating temperature and interference. The magnitude of interference fit of steel joint was predicted. The magnitude of the interference fit of the steel joint was predicted. A polynomial model, exponential model and Gaussian model were adopted to predict the sectional deformation so as to compare and analyze the predictive performance of a BP neural network, among which the polynomial model was used to predict the magnitude of the interference fit. Through a comparative analysis of the fitting residual (RE) and sum of squares of the error (SSE), it can be known that a BP neural network has good prediction accuracy. The predicted results showed that the error of the prediction model increases with the increase of the heating temperature in the prediction model of the steel node interference and related factors. When the cooling water velocity hit 0.038 m/s, the prediction accuracy was the highest. The prediction error increases with the increase or decrease of the velocity. Especially when the velocity increases, the trend of error increasing became more obvious. The analysis shows that this method has better prediction accuracy.

On the Utilization of an Ensemble of Meta-Heuristics for Simulating Energy Consumption in Buildings

Predicting energy consumption has been a substantial topic because of its ability to lessen energy wastage and establish an acceptable overall operational efficiency. Thus, this research aims at creating a meta-heuristic-based method for autonomous simulation of heating and cooling loads of buildings. The developed method is envisioned on two tiers, whereas the first tier encompasses the use of a set of meta-heuristic algorithms to amplify the exploration and exploitation of Elman neural network through both parametric and structural learning. In this regard, ten meta-heuristic were utilized, namely differential evolution, particle swarm optimization, invasive weed optimization, teaching-learning optimization, ant colony optimization, grey wolf optimization, grasshopper optimization, moth-flame optimization, antlion optimization, and arithmetic optimization. The second tier is designated for evaluating the meta-heuristic-based models through performance evaluation and statistical comparisons. Besides, an integrative ranking of the models is achieved using average ranking algorithm.

A New Prediction Method for the Complete Characteristic Curves of Centrifugal Pumps

Complete pump characteristics (CPCs) are the key for establishing pump boundary conditions and simulating hydraulic transients. However, they are not normally available from manufacturers, making pump station design difficult to carry out. To solve this issue, a novel method considering the inherent operating characteristics of the centrifugal pump is therefore proposed to predict the CPCs. First, depending on the Euler equations and the velocity triangles at the pump impeller, a mathematical model describing the complete characteristics of a centrifugal pump is deduced. Then, based on multiple measured CPCs, the nonlinear functional relationship between the characteristic parameters of the characteristic operating points (COPs) and the specific speed is established. Finally, by combining the mathematical model with the nonlinear relationship, the CPCs for a given specific speed are successfully predicted. A case study shows that the predicted CPCs are basically consistent with the measured data, showing a high prediction accuracy. For a pump-failure water hammer, the simulated results using the predicted CPCs are close to that using the measured data with a small deviation. This method is easy to program and the prediction accuracy meets the requirements for hydraulic transient simulations, providing important data support for engineering design.

Prediction of Slag Characteristics Based on Artificial Neural Network for Molten Gasification of Hazardous Wastes

Molten gasification is considered as a promising technology for the processing and safe disposal of hazardous wastes. During this process, the organic components are completely converted while the hazardous materials are safely embedded in slag via the fusion-solidification-vitrification transformation. Ideally, the slag should be glassy with low viscosity to ensure the effective immobilization and steady discharge of hazardous materials. However, it is very difficult to predict the characteristics of slag using existing empirical equations or conventional mathematical methods, due to the complex non-linear relationship among the phase transformation, vitrification transition and chemical composition of slag. Equipped with a strong nonlinear mapping ability, an artificial neural network may be able to predict the properties of slags if a large amount of data is available for training. In this work, over 10,000 experimental data points were used to train and develop a slag classification model (glassy vs. non-glassy) based on a neural network. The optimal structure of the neural network was figured out and validated. The results suggest that the classification accuracy for the independent test samples reached 93.3%. Using 1 and 0 as model inputs to represent mildly reducing and inert atmospheres, a double hidden layer structure in the neural network enabled the accurate classification of slags under various atmospheres. Furthermore, the neural network for the prediction of glassy slag viscosity was optimized; it featured a double hidden layer structure. Under a mildly reducing atmosphere, the absolute error from the independent test data was generally within 4 Pa·s. By adding a gas atmosphere into the input of the neural network using a simple normalization method, a multi-atmosphere slag viscosity prediction model was developed. Said model is much more accurate than its counterpart that does not consider the effect of the atmosphere. In summary, the artificial neural network proved to be an effective approach to predicting the slag properties under different atmospheres. The data-driven models developed in this work are expected to facilitate the commercial deployment of molten gasification technology.

Effects of a Combination Impeller on the Flow Field and External Performance of an Aero-Fuel Centrifugal Pump

Aero-fuel centrifugal pumps are important power plants in aero-engines. Unlike most of the existing centrifugal pumps, a combination impeller is integrated with the pump to improve performance. First, the critical geometrical parameters of the combination impeller and volute are given. Then, the effects of the combination impeller on the flow characteristics of the impeller and volute are clarified by comparing simulation results with that of the conventional impeller, where the effectiveness of the selected numerical method is validated by an acceptable agreement between simulation and experiment. Finally, the experiment is set to test the external performance of the studied pump. A significant feature of this study is that the flow characteristics are significantly ameliorated by reducing the flow losses that emerged in the impeller inlet, impeller outlet, and volute tongue. Correspondingly, the head and efficiency of a combination impeller are higher with comparison to a conventional impeller. Consequently, it is a promising approach in ameliorating the flow field and improving external performance by applying a combination impeller to an aero-fuel centrifugal pump.