Accurate Temperature Prediction Research Articles

Sea surface temperature prediction by stacked generalization ensemble of deep learning

Accurate sea surface temperature (SST) prediction is of great significance for fishery farming, marine ecological protection, and planning of maritime activities. In this paper, the stacked generalization ensemble is demonstrated for SST prediction to improve every single deep-learning model. Long-term high-resolution satellite-derived SST is used with the sub-regions of the Taiwan Strait and East China Sea taken as the study area. We select the Multilayer Perceptron, Long short-term memory (LSTM), Convolutional Neural Networks (CNN), CNN-LSTM as individual learners, and Convolutional LSTM as the meta-learner. The individual learners are trained and validated on the retained data subset I, while the meta-learner is trained and validated by constructing the samples with the predictions of validated individual learners on the retained data subset II. The two types of models are evaluated on the same test dataset with root-mean-square error and coefficient of efficiency as the scoring criteria. We find that the meta-model outperforms any individual model and other baselines for the one-day-/three-day-ahead forecasts in the Taiwan Strait and one-day-/three-day-/five-day-ahead predictions in the East China Sea. Furthermore, when the lead time is 1 day and 3 days, the meta-model has a better spatial distribution of prediction metrics across all grid points in the Taiwan Strait sub-area. For the East China Sea sub-region, the meta-model advantage is extended to the lead time of 5 days. Probably due to the higher quality of offshore satellite data, the prediction ability enhancement of the stacked ensemble applied in the East China Sea is better than that in the Taiwan Strait. The better-performing meta-model prediction suggests that the stacked generalization ensemble is encouraging and promising for improving the short-term prediction of the daily SST field.

Development of machine learning models for estimation of daily evaporation and mean temperature: a case study in New Delhi, India

ABSTRACT Accurate prediction of pan evaporation and mean temperature is crucial for effective water resources management, influencing the hydrological cycle and impacting water availability. This study focused on New Delhi's semi-arid climate, data spanning 31 years (1990–2020) were used to predict these variables using advanced algorithms such as Bagging, Random Subspace (RSS), M5P, and REPTree. The models were rigorously evaluated using 10 performance metrics, including correlation coefficient, mean absolute error (MAE), and Nash–Sutcliffe Efficiency (NSE) model coefficient. The Bagging model emerged as the best model with performance indices values as r, MAE, RMSE, RAE, RRSE, MBE NSE, d, KGE, and MAPE as 0.86, 0.76, 1.43, 32.70, 49.44, 0.03, 0.85, 0.96, 0.90, and 22.0, respectively, during model testing phase for pan evaporation prediction. In predicting mean temperature, the Bagging model reported the best results with performance indices values as r, MAE, RMSE, RAE, RRSE, MBE NSE, d, KGE, and MAPE as 0.86, 0.76, 1.43, 32.70, 49.44, 0.03, 0.85, 0.96, 0.90 and 22.0, respectively, during the model testing phase. These findings offer valuable insights for enhancing relative humidity prediction models in diverse climatic conditions. The Bagging model's robust performance underscores its potential application in water resource management.

Synergizing Transfer Learning and Multi-Agent Systems for Thermal Parametrization in Induction Traction Motors

Maintaining optimal temperatures in the critical parts of an induction traction motor is crucial for railway propulsion systems. A reduced-order lumped-parameter thermal network (LPTN) model enables computably inexpensive, accurate temperature estimation; however, it requires empirically based parameter estimation exercises. The calibration process is typically performed in labs in a controlled experimental setting, which is associated with a lot of supervised human efforts. However, the exploration of machine learning (ML) techniques in varied domains has enabled the model parameterization in the drive system outside the laboratory settings. This paper presents an innovative use of a multi-agent reinforcement learning (MARL) approach for the parametrization of an LPTN model. First, a set of reinforcement learning agents are trained to estimate the optimized thermal parameters using the simulated data in several driving cycles (DCs). The selection of a reinforcement learning agent and the level of neurons in the RL model is made based on variability of the driving cycle data. Furthermore, transfer learning is performed on a new driving cycle data collected on the measurement setup. Statistical analysis and clustering techniques are proposed for the selection of an RL agent that has been pre-trained on the historical data. It is established that by synergizing within reinforcement learning techniques, it is possible to refine and adjust the RL learning models to effectively capture the complexities of thermal dynamics. The proposed MARL framework shows its capability to accurately reflect the motor’s thermal behavior under various driving conditions. The transfer learning usage in the proposed approach could yield significant improvement in the accuracy of temperature prediction in the new driving cycles data. This approach is proposed with the aim of developing more adaptive and efficient thermal management strategies for railway propulsion systems.

A proposed hybrid model of ANN and KNN for solar cell defects detection and temperature prediction using fuzzy image segmentation

This paper presents a novel hybrid model employing Artificial Neural Networks (ANN) and Mathematical Morphology (MM) for the effective detection of defects in solar cells. Focusing on issues such as broken corners and black edges caused by environmental factors like broken glass cover, dust, and temperature variations. This study utilizes a hybrid model of ANN and K-Nearest Neighbor (KNN) for temperature prediction. This hybrid approach leverages the strengths of both models, potentially opening up new avenues for improved accuracy in temperature forecasting, which is critical for solar energy applications. The significance lies in the interconnectedness of temperature fluctuations and solar cell efficiency, leading to defects. The proposed model aims to predict temperatures accurately, providing insights into potential solar cell efficiency problems. Subsequently, this work studies the transitions to defect detection using Fuzzy C-Means (FCM) clustering and MM techniques. The hybrid model demonstrates accurate temperature prediction with Mean Absolute Percentage Error (MAPE) values of 0.92 %, 0.72 %, and 1.3 % for average, maximum, and minimum temperatures, respectively. The defect detection process yields a detection accuracy (CR) of 96 % and sensitivity of detection (SD) of 89 %. This work is validated compared to the literature work done and by using K-fold cross validation technique. The proposed work emphasizes the improvement in defect detection accuracy and the overall quality enhancement of solar cells.

A Combustion Kinetic Model Based on the Fast Chemistry Assumption: Application to Non-Sooty and Sooty Laminar Diffusion Flames

ABSTRACT In this article, a combustion kinetic model based on the fast chemistry (FC) assumption is proposed to compute the finite reaction rate of gaseous species in laminar diffusion flames. The algorithm developed for this purpose is numerically robust and relies on the adiabatic flame temperature to limit the fuel consumption rate. While there is no FC model specifically developed for laminar flames, the Eddy Dissipation Model (EDM) correction scheme for laminar regimes is taken as the reference model. The proposed model addresses the shortcomings of the reference model, i.e. the sensitivity to the transport time scale and the cell size and the need to calibrate a numerical constant repeatedly. The new model formulation does not rely on any numerical constant and completely removes the dependency on the cell size and the species transport time scale, as shown for multiple counterflow and axisymmetric laminar diffusion flames. Furthermore, the proposed model provides consistent temperature and species mass fraction predictions across the mentioned flames. Finally, a sooty axisymmetric flame is simulated using the mentioned models to investigate the numerical interaction between combustion and soot chemistry. The results show that the reference model requires recalibration to capture the temperature and soot volume fraction distributions while the proposed model yields reasonably accurate temperature and soot predictions.

[formula omitted]-Functions for fields of series- and parallel-connected boreholes with variable fluid mass flow rate and reversible flow direction

A new methodology is developed for the calculation of g-functions for the simulation of geothermal bore fields in non-stationary conditions. The g-functions are able to represent the variations of fluid mass flow rate and reversible flow direction, and model the effect of these variations on the long-term ground temperature response. The thermal model is constructed by coupling an axially-discretized finite line source solution for the ground heat transfer, a thermal resistance circuit model for the interior of the boreholes, as well as connectivity relations between parallel- and series-connected boreholes. Simulation experiments show that the new g-functions are required for the accurate prediction of fluid temperatures in series-connected boreholes with variable mass flow rate and reversible flow direction. A simulation of a borehole thermal energy storage system of 144 boreholes over a period of 20 years shows a maximum absolute error of 0.65 °C. The new g-functions thus extend the simulation capabilities of g-functions to borehole thermal energy storage systems.

A computationally efficient thermo-mechanical model with temporal acceleration for prediction of residual stresses and deformations in WAAM

ABSTRACT The present paper introduces a novel temporal acceleration strategy for computationally efficient prediction of residual stresses and deformations in wire arc additive manufacturing (WAAM) components. It employs a semi-analytical approach, dividing temperature into the analytical and the complementary fields. The analytical field is obtained by a closed-form solution, and the complementary field is employed to solve the boundary conditions. A temporal acceleration factor for the heating period is applied. Meanwhile, the diffusion time in the analytical field is manipulated to guarantee accurate temperature prediction. Validation via WAAM experiments, including both a thin-wall structure and a practical engineering component with characteristic dimensions on the order of metres, indicates that the predicted stresses is 50 MPa lower than the experimental values. Moreover, the discrepancy of between the predicted deformations and the experimental measurements is less than 10%, demonstrating reasonable accuracy can be achieved with attractive computational efficiency.

Modeling and experimental study of air-cooled finned-tube condenser using coupled distribution parameter method and air-side temperature distribution

The air-cooled finned tube condenser (ACFTC) plays a crucial role in refrigeration systems, profoundly influencing overall operational efficiency. Its intricate structure, connectivity, and phase change characteristics pose challenges for numerical simulation modeling. While traditional distribution parameter methods effectively capture flow and heat transfer within tube mediums, they overlook ACFTC’s air-side temperature distribution, resulting in discrepancies between simulated and experimental outcomes. This study enhances existing distribution parameter models by incorporating ACFTC’s air-side temperature distribution characteristics, yielding an improved predictive model for transient operational behavior of both Z-shaped and U-shaped configurations. Dynamic experiments and model analyses were performed on the Z-shaped configuration using a cold storage experimental setup. Results demonstrate the improved model’s accurate prediction of tubing medium outlet temperature and pressure, with maximum deviations of approximately 3 °C and 0.01 MPa, respectively. Notably, the improved model exhibits around a 30 % enhancement in predictive accuracy compared to the existing models for both Z-shaped and U-shaped ACFTCs.

In-flight anti-icing simulation of electrothermal ice protection systems with inhomogeneous thermal boundary condition

Conventional anti-icing computational solvers calculate the convective heat transfer coefficient using a homogeneous thermal boundary condition, assuming a constant temperature across the surface. However, this approach can lead to inaccuracies in regions with significant temperature variations. To address this limitation, the present study proposes a novel approach that utilizes an inhomogeneous thermal boundary condition, which updates the convective heat transfer coefficient based on the transient wall temperature distribution. We developed a unified finite volume framework that efficiently integrates various in-house solvers, including a compressible Navier-Stokes-Fourier (NSF) airflow solver, a Eulerian droplet impingement solver, a PDE-based ice solver, and a multilayer heat conduction solver. Thermal interaction between the different solvers was modeled using the conjugate heat transfer method. The effects of updating airflow on anti-icing results were investigated by comparing the results obtained by coupled and decoupled solvers. While the decoupled solver computes airflow once and remains unchanged, the coupled solver updates the airflow during the anti-icing simulation. Our findings show that the decoupled solver has the highest deviations from the coupled solver in evaporative anti-icing regimes, where dry regions form within the protection limits. Using coupled solvers in designing ice protection systems in evaporative regimes improves temperature prediction accuracy, enabling designers to reduce safety factors and save energy.

Artificial intelligence applications for accurate geothermal temperature prediction in the lower Friulian Plain (north-eastern Italy)

Geothermal energy as a sustainable and clean energy source depends on the accurate estimation of reservoir temperatures. Understanding aquifer temperatures is crucial for optimizing low-enthalpy geothermal system exploitation. Advances in predictive algorithms can improve geothermal efficiency, while conventional methods of indirect temperature measurement and assumptions in geochemical analysis lead to uncertainties. As a solution, this study presents a comprehensive evaluation of six machine learning algorithms including eXtreme gradient boosting (XGBoost), decision tree, generalized regression neural network, extreme randomized trees, radial basis function, and elastic net. We employed essential performance metrics including coefficient of determination (R2) score, root mean square error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and variance accounted for (VAF) to elucidate their predictive accuracy and generalization potential in the lower Friulian Plain (north-eastern Italy) where a geothermal reservoir is present. Among the algorithms scrutinized, XGBoost emerges as a predictive exemplar, achieving a remarkable R2 score of 0.9930 on the test dataset, with consistently low RMSE of 0.788, MAE of 0.587, MAPE of 1.909, and high VAF of 99.30, reaffirming its exceptional precision and robustness. It is worth noting that the other four models show slightly weaker performance than XGBoost, while elastic net shows moderate predictive power, which illustrates the complexity of the database. The Wilcoxon signed-rank test confirmed the superior performance of XGBoost in estimating geothermal temperatures compared to other algorithms, with statistical evidence supporting its precision and reliability. A Monte Carlo simulation for uncertainty analysis underlined the importance of model selection, accuracy and uncertainty management in the planning of geothermal projects in the lower Friulian Plain. A sensitivity analysis was performed to identify the main factors influencing the temperature prediction. Among the parameters considered, bicarbonate the highest significance at 0.51, which is essential for accurate temperature prediction because of its buffering capacity which directly influences water's thermal properties. Magnesium and electrical conductivity each contribute with 0.11, also play significant roles due to their impact on the water's heat retention and distribution capabilities. Water depth, with a value of 0.08, also has a significant influence on the temperature profiles in prediction models. In summary, the accurate prediction of XGBoost for the temperature of aquifer in carbonate reservoirs in the lower Friulian Plain, underline its value for optimizing geothermal resources and highlight most important influences on temperature.

Predicting hourly heating load in residential buildings using a hybrid SSA–CNN–SVM approach

This study proposes a hybrid prediction model using sparrow search algorithm (SSA) to optimize the convolutional neural network (CNN) and support vector machine (SVM), in order to perform accurate prediction of secondary supply temperature (Ts2). The historical operation data of Weifang residential building thermal station was adopted and reasonable data preprocessing was performed to suppress the interference of abnormal data. The input variables of the prediction model were screened using the correlation analysis method, taking the influence of the hysteresis effect into consideration. The SSA–CNN–SVM model was then developed for prediction. The performance of the model was evaluated by the root mean square error, mean absolute error, mean absolute percentage error (MAPE), and absolute value of relative error of each time step. The results obtained demonstrated that the SSA–CNN–SVM model has high prediction accuracy. The MAPE values of the two heat exchange stations were between 2.28 % and 2.4 %. The indoor temperature significantly affected the prediction accuracy of Ts2. After the introduction of the indoor temperature, the MAPE of the predicted values of the hybrid model was reduced by 0.35 %. The maximum reduction in MAPE of the SSA–CNN–SVM model was 1.5 % compared with other prediction models.

Patient-specific temperature distribution prediction in laser interstitial thermal therapy: single-irradiation data-driven method

Laser interstitial thermal therapy (LITT) is popular for treating brain tumours and epilepsy. The strict control of tissue thermal damage extent is crucial for LITT. Temperature prediction is useful for predicting thermal damage extent. Accurately predicting in vivo brain tissue temperature is challenging due to the temperature dependence and the individual variations in tissue properties. Considering these factors is essential for improving the temperature prediction accuracy. Objective. To present a method for predicting patient-specific tissue temperature distribution within a target lesion area in the brain during LITT. Approach. A magnetic resonance temperature imaging (MRTI) data-driven estimation model was constructed and combined with a modified Pennes bioheat transfer equation (PBHE) to predict patient-specific temperature distribution. In the PBHE for temperature prediction, the individual specificity and temperature dependence of thermal tissue properties and blood perfusion, as well as the individual specificity of optical tissue properties were considered. Only MRTI data during one laser irradiation were required in the method. This enables the prediction of patient-specific temperature distribution and the resulting thermal damage region for subsequent ablations. Main results. Patient-specific temperature prediction was evaluated based on clinical data acquired during LITT in the brain, using intraoperative MRTI data as the reference standard. Our method significantly improved the prediction performance of temperature distribution and thermal damage region. The average root mean square error was decreased by 69.54%, the average intraclass correlation coefficient was increased by 37.5%, the average Dice similarity coefficient was increased by 43.14% for thermal damage region prediction. Significance. The proposed method can predict temperature distribution and thermal damage region at an individual patient level during LITT, providing a promising approach to assist in patient-specific treatment planning for LITT in the brain.

Reservoir temperature prediction based on characterization of water chemistry data—case study of western Anatolia, Turkey

Reservoir temperature estimation is crucial for geothermal studies, but traditional methods are complex and uncertain. To address this, we collected 83 sets of water chemistry and reservoir temperature data and applied four machine learning algorithms. These models considered various input factors and underwent data preprocessing steps like null value imputation, normalization, and Pearson coefficient calculation. Cross-validation addressed data volume issues, and performance metrics were used for model evaluation. The results revealed that our machine learning models outperformed traditional fluid geothermometers. All machine learning models surpassed traditional methods. The XGBoost model, based on the F-3 combination, demonstrated the best prediction accuracy with an R2 of 0.9732, while the Bayesian ridge regression model using the F-4 combination had the lowest performance with an R2 of 0.8302. This study highlights the potential of machine learning for accurate reservoir temperature prediction, offering geothermal professionals a reliable tool for model selection and advancing our understanding of geothermal resources.

Neural network for enhancement of end milling processes through accurate prediction of temperature in the cutting zone

Neural network for enhancement of end milling processes through accurate prediction of temperature in the cutting zone

Fast thermodynamic simulation of electrical control cabinets via thermal resistor networks

A fine-grid Computational Fluid Dynamics (CFD) simulation without thermo-electric coupling provides an accurate prediction of temperatures inside low-voltage control cabinets. But due to its high computational demand and complexity, the approach is not eligible for a widespread use in pre-production design. This study presents a lightweight and robust model based on the thermal-electrical analogy and Kirchhoff’s current law. The thermal resistor network is created for each control cabinet individually. An isothermal coarse-grid CFD simulation is used to approximate flow resistances. Convective resistances are calculated with Nusselt number correlations. The approach was tested on 120 randomly generated 2D control cabinet configurations. Component temperature errors were evaluated based on a normal distribution approximation. For moderate temperature rises up to 20 K the model had a mean error of 0.38 K and a standard error of 1.51 K compared to the reference fine-grid CFD model and was up to 12.8 times faster.

Exploring Machine Learning Techniques for Accurate Prediction of Methane Hydrate Formation Temperature in Brine: A Comparative Study

AbstractAccurate estimation of formation conditions plays a pivotal role in effectively managing various processes related to hydrates, including flow assurance, deep-water drilling, and hydrate-based technology development. The formation temperature of methane hydrates in the presence of brine greatly affects the efficacy and accuracy of these processes. This work presents a comprehensive and novel comparative analysis of nine distinct machine learning models for accurate prediction of formation temperatures of methane hydrate. This study investigated the application of major machine learning (ML) algorithms including multiple linear regression (MLR), long short-term memory (LSTM), radial basis function (RBF), support vector machine (SVM), artificial neural network (ANN), gradient boosting regression (GBR), gradient process regression (GPR), random forest (RF), and K-nearest neighbor (KNN). The model accuracy was validated against a large dataset comprising of over 1000 data points with diverse range of salt concentrations. In this regard, model accuracies were compared using several metrics including R2, ARD, and AARD. The experimental results exhibited KNN algorithm to be fast-converging, accurate, and consistent over the entire range of data points with an R2 score of 0.975 and AARD of 0.385%. The results enable efficient and accurate temperature estimation with ML algorithms for multiple hydrate-related processes.

Cracking the code: Spatial heterogeneity as the missing piece for modeling granular fluidized bed drying

Pharmaceutical twin-screw wet granulation is a multifaceted and intricate process pivotal to drug product development. Accurate modeling of this process is indispensable for optimizing manufacturing parameters and ensuring product quality. The fluid bed dryer, an integral component of this granulation process, significantly influences the granular critical quality attributes. This study builds upon prior research by integrating experimental findings on granule segregation during fluid bed drying into an existing compartmental model, enhancing its predictive capabilities. An additional model layer on granule segregation behavior is composed and integrated into the existing model structure in this study. The added model compartment describes probability distributions on the vertical position of granules within each granule size class considered. To beware of overfitting, predictions of both the moisture content after drying and the granule bed temperature throughout drying are discussed in this study relative to experimental data from earlier published studies. These independent analyses demonstrated a marked improvement in prediction accuracy compared to earlier published model structures. The refined model accurately predicts the residual moisture content after drying for an untrained formulation. Moreover, it simultaneously makes accurate predictions of the granular bed temperature, which emboldens its structural correctness. This advancement makes it a powerful tool for predicting the behavior of the pharmaceutical fluid bed drying, which holds significant promise to facilitate pharmaceutical product development.

Chasing the heat: Unraveling urban hyperlocal air temperature mapping with mobile sensing and machine learning

Many cities face unprecedented high temperatures with increasing extreme events. Heatwaves pose significant health risks, including cardiovascular diseases, heatstroke, and dehydration. Mapping urban near-surface air temperature (Tair) is crucial for understanding thermal exposure and addressing climate change. Previous studies relied on satellite-derived land surface temperature (LST) and stationary monitoring, but high spaio-temporal Tair mapping is still a challenge. This study optimized a mobile sensing scheme using an electric bicycle platform with environmental and image sensors, and deep learning captured local-scale urban factors. A spatio-temporal data fusion model that consisted of three parts, temporal trend extraction, locality analysis, and neighborhood effect analysis, generated hyperlocal Tair maps. The Results from Beijing demonstrated the effectiveness of the framework, achieving the lowest MAE of 0.02 °C. Optimized data collection and the new model achieved accurate temperature predictions and thermal exposure assessment. Efficiency enhanced sensing strategy was also proposed. The study highlights local-scale factors and spatio-temporal dependencies in addressing heatwaves and climate change impacts in urban areas.

Comparison of time series temperature prediction with auto-regressive integrated moving average and recurrent neural network

The region of Beni Mellal, Morocco is heavily dependent on the agricultural sector as its primary source of income. Accurate temperature prediction in agriculture has many benefits including improved crop planning, reduced crop damage, optimized irrigation systems and more sustainable agricultural practices. By having a better understanding of the expected temperature patterns, farmers can make informed decisions on planting schedules, protect crops from extreme temperature events, and use resources more efficiently. The lack of data-driven studies in agriculture impedes the digitalization of farming and the advancement of accurate long-term temperature prediction models. This underscores the significance of research to identify the optimal machine learning models for that purpose. A 22-year time series dataset (2000-2022) is used in the study. The machine-learning model auto-regressive integrated moving average (ARIMA) and deep learning models simple recurrent neural network (SimpleRNN), gated recurrent unit (GRU), and long short-term memory (LSTM) were applied to the time series. The results are evaluated based on the mean absolute error (MAE). The findings indicate that the deep learning models outperformed the machine-learning model, with the GRU model achieving the lowest MAE.

A new LSTNet-based temperature prediction model for permanent magnet

Permanent magnet synchronous motors (PMSMs) can effectively protect against demagnetization by accurate permanent magnet (PM) temperature prediction; nevertheless, due to the nonlinear properties and intricate internal structure of PMSMs, accurate PM temperature prediction methods still encounter difficulties. This paper proposes a new PM temperature prediction model (LSTNet-Improved) based on long- and short-term time series network (LSTNet) to increase the prediction accuracy of PM temperature. By adding a multi-scale convolutional (MSC) layer to the convolutional neural network (CNN) layer of LSTNet, the short-term detail-dependent information between PMSM variables can be obtained by the model, and adds a CNN skip (CNN-skip) layer in place of the gate recurrent unite skip (GRU-skip) layer to find the long- and short-term local repetitive patterns between the variables, which yields additional feature information. Furthermore, a nonlinear bias term (MNB) is added to the original multi-head attention layer in order to solve the complex nonlinear relationship between multivariate time series. These layers are applied after the CNN-skip layer and gated recurrent unite (GRU) layers, respectively. This allows the model to become more robust and automatically learn to the complex nonlinear relationship between the sequences, focus on the important feature information, and lessen the impact of redundant information on the PMs’ temperature prediction task. The experimental results show that the goodness of fit (R2), root mean square error (RMSE) and mean absolute error (MAE) of the proposed model are 99.93%, 14.55% and 9.45%, respectively, and compared with LSTNet, the R2 is improved by 0.09%, the RMSE and MAE are reduced by 7.03% and 6.51%. These results confirm that the model can correctly predict the PM temperature, efficiently extract both long- and short-term patterns among the variables of the PMSM, and effectively focus on the key feature information in the intricate nonlinear series.