| 1. |
- Aslanidou, Ioanna, et al.
(författare)
-
Comparison of machine learning approaches for spectroscopy applications
- 2022
-
Ingår i: Proceedings of the 63rd International Conference of Scandinavian Simulation Society. - : Linköping University Electronic Press. - 9789179295455 ; , s. 80-85
-
Konferensbidrag (refereegranskat)abstract
- In energy production the characterization of the fuel is a key aspect for modelling and optimizing the operation of a power plant. Near-infrared spectroscopy is a wellestablished method for characterization of different fuels and is widely used both in laboratory environments and in power plants for real-time results. It can provide a fast and accurate estimate of key parameters of the fuel, which for the case of biomass can include moisture content, heating value, and ash content. These instruments provide a chemical fingerprint of the samples and require a calibration model to relate that to the parameters of interest.A near-infrared spectrometer can provide point data whereas a hyperspectral imaging camera allows the simultaneous acquisition of spatial and spectral information from an object. As a result, an installation above a conveyor belt can provide a distribution of the spectral data on a plane. This results in a large amount of data that is difficult to handle with traditional statistical analysis. Furthermore, storage of the data becomes a key issue, therefore a model to predict the parameters of interest should be able to be updated continuously in an automated way. This makes hyperspectral imaging data a prime candidate for the application of machine learning techniques. This paper discusses the modelling approach for hyperspectral imaging, focusing on data analysis and assessment of machine learning approaches for the development of calibration models.
|
|
| 2. |
- Diamantidou, Dimitra-Eirini, 1995-, et al.
(författare)
-
Navigating Technological Risks : An Uncertainty Analysis of Powertrain Technology in Hybrid-Electric Commuter Aircraft
- 2024
-
Ingår i: Proceedings of the ASME Turbo Expo 2024 Turbomachinery Technical Conference and Exposition.
-
Konferensbidrag (refereegranskat)abstract
- This study addresses the uncertainties in hybrid-electric powertrain technology for a 19-passenger commuter aircraft, focusing on two future Entry-Into-Service timeframes: 2030 and 2040. The methodology is split into a preliminary optimization of aircraft design based on nominal technology scenarios followed by Monte Carlo simulations to investigate the impact of diverse technology projections and distribution types. Advanced surrogate modeling techniques, leveraging deep neural networks trained on a dataset from an aircraft design framework, are employed.Key outcomes from this work reveal a marked increase in computational efficiency, with a speed-up factor of approximately 500 times when utilizing surrogate models. The results indicate that the 2040 EIS scenario could achieve larger reductions in fuel and total energy consumption—20.4% and 15.8% respectively—relative to the 2030 scenario, but with higher uncertainty. Across all scenarios examined, the hybrid-electric model showcased superior performance compared to its conventional counterpart. The battery specific energy density is proved to be a critical parameter of the aircraft's performance across both timeframes. The findings emphasize the importance of continuous innovation in battery and motor technologies to target towards greater system-level efficiency and reduced environmental impact.
|
|
| 3. |
- Helmryd Grosfilley, E., et al.
(författare)
-
Investigation of Machine Learning Regression Techniques to Predict Critical Heat Flux over a Large Parameter Space
- 2024
-
Ingår i: Nuclear Technology. - : Taylor & Francis Inc. - 0029-5450 .- 1943-7471.
-
Tidskriftsartikel (refereegranskat)abstract
- A unifying and accurate model to predict Critical Heat Flux (CHF) over a wide range of conditions has been elusive since wall boiling research emerged. With the release of the data utilized in the development of the 2006 Groeneveld CHF lookup table (LUT), by far the most extensive public CHF database available to date (nearly 25000 data points), development of data-driven predictions models over a large parameter space in simple geometry (vertical, uniformly heated round tubes) can be performed. Furthermore, the popularization of machine learning techniques to solve regression problems has led to more advanced tools for analyzing large and complex databases. This work compares three machine learning algorithms to predict the entire LUT CHF test database. For each selected regression algorithm (ν-Support vector, Gaussian process, and neural network), an optimized hyperparameter set is fitted. The best-performing algorithm is the neural network, which can achieve a standard deviation of the predicted/measured factor of 12.3%, three times lower than the LUT. In comparison, the Gaussian process regression and the ν-Support vector regression achieve a standard deviation of 17.7%, about two times lower than the LUT. All considered algorithms hence significantly outperform the LUT prediction performance. The neural network model and training methodology are designed to prevent overfitting, which is confirmed by data analysis of the predictions. Finally, a feasibility study of transfer learning is presented and future development directions (including uncertainty quantification) are discussed.
|
|
| 4. |
- Kyprianidis, Konstantinos, et al.
(författare)
-
On-line Powerplant Control using Near-InfraRed Spectroscopy : OPtiC-NIRS, REPORT 2021:746
- 2021
-
Rapport (övrigt vetenskapligt/konstnärligt)abstract
- Near InfraRed Spectroscopy (NIRS) offers rapid on-line analysis of biomass feedstocks and can be utilized for process control of biomass- based combined heat and power plants. Within the OPtiC-NIRS project we have carried out a full-scale on-site testing of different NIRS for online powerplant control at the facilities of Mälarenergi and Eskilstuna Strängnäs Energi och Miljö. The project has been focused on developing and testing robust NIRS soft-sensors for fuel higher heating value and composition (incl. moisture, components such as recycle wood and glass, different type of plastics and ash) and combining them with dynamic models for on-line feed-forward process monitoring and control. Expected benefits include reduced risk of agglomeration and pollutant emissions formation as well as improved production control. A longer-term potential and ambition is to be able to identify the fossil content in municipal waste fuel, which can hopefully be addressed in a follow-up study.
|
|
| 5. |
- Skvaril, Jan, 1982-, et al.
(författare)
-
Application of single-point and hyperspectral imaging near-infrared sensors and machine learning algorithms for real-time biomass characterization
- 2019
-
Konferensbidrag (refereegranskat)abstract
- Biomass is typically a material with highly variable properties making its use in industrial combustion processes challenging due to requirements on the steady operation. Property such as moisture content has an impact on fuel ignition characteristics and heat release from the biomass. Ash content negatively influences fluidization of the boiler bed and after-burning of small fuel particles, by forming an impermeable layer on the surface resulting in incomplete combustion and formation of harmful emissions.The large variability of the properties thus creates undesired process instabilities which need to be addressed in a timely manner by appropriate operational/regulatory measures adjusting e.g. fluidization velocity, distribution of combustion air, under-pressure in the furnace etc. Consequently, there is a need for the implementation of sensors able to measure the properties of interest in real-time. In our previous studies, we demonstrated the ability of a single-point near-infrared sensor to measure fuel properties in real-time in a laboratory environment. However, we found that there is limited representativeness of the single-point measurements as also a cross-sectional variation of the fuel properties on the conveyor belt was apparent.Therefore, the implementation of a sensor able to measure also a spatial distribution of the material in the biomass stream is suggested. Literature review shows that it can be achieved by the implementation of a near-infrared hyperspectral imaging camera.The aim of the work is to present research activities at the Future Energy Center, Mälardalen University leading towards the installation of a) single-point and b) hyperspectral imaging near-infrared sensors for real-time moisture and ash content measurements. The study further presents the concept of NIR sensors integration for process optimization and the introduction of new advanced control concepts for steam boilers.
|
|
| 6. |
- Soibam, Jerol, et al.
(författare)
-
A Data-Driven Approach for the Prediction of Subcooled Boiling Heat Transfer
- 2020
-
Ingår i: Proceedings of The 61st SIMS Conference on Simulation and Modelling SIMS 2020. - : Linköping University Electronic Press. ; , s. 435-442
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)abstract
- In subcooled flow boiling, heat transfer mechanism involves phase change between liquid phase to the vapour phase. During this phase change, a large amount of energy is transferred, and it is one of the most effective heat transfer methods. Subcooled boiling heat transfer is an attractive trend for industrial applications such as cooling electronic components, supercomputers, nuclear industry, etc. Due to its wide variety of applications for thermal management, there is an increasing demand for a faster and more accurate way of modelling. In this work, a supervised deep neural network has been implemented to study the boiling heat transfer in subcooled flow boiling heat transfer. The proposed method considers the near local flow behaviour to predict wall temperature and void fraction of a sub-cooled mini-channel. The input of the network consists of pressure gradients, momentum convection, energy con- vection, turbulent viscosity, liquid and gas velocities, and surface information. The output of the model is based on the quantities of interest in a boiling system i.e. wall temperature and void fraction. The network is trained from the results obtained from numerical simulations, and the model is used to reproduce the quantities of interest for interpolation and extrapolation datasets. To create an agile and robust deep neural network model, state-of-the-art methods have been implemented in the network to avoid the overfitting issue of the model. The results obtained from the deep neural network model shows a good agreement with the numerical data, the model has a maximum relative error of 0.5 % while predicting the temperature field, and for void fraction, it has approximately 5 % relative error in interpolation data and a maximum 10 % relative error for the extrapolation datasets.
|
|
| 7. |
- Soibam, Jerol, et al.
(författare)
-
Application of deep learning for segmentation of bubble dynamics in subcooled boiling
- 2023
-
Ingår i: International Journal of Multiphase Flow. - 0301-9322 .- 1879-3533. ; 169
-
Tidskriftsartikel (refereegranskat)abstract
- The present work focuses on designing a robust deep-learning model to track bubble dynamics in a vertical rectangular mini-channel. The rectangular mini-channel is heated from one side with a constant heat flux, resulting in the creation of bubbles. Images of the bubbles are recorded using a high-speed camera, which serve as the input data for the deep learning model. The raw image data acquired from the high-speed camera is inherently noisy due to the presence of shadows, reflections, background noise, and chaotic bubbles. The objective is to extract the mask of the bubble given all these challenging factors. Transfer learning is adopted to eliminate the need for a large dataset to train the deep learning model and also to reduce computational costs. The trained model is then validated against the validation datasets, demonstrating an accuracy of 98% while detecting the bubbles. The model is then evaluated on different experimental conditions, such as lighting, background, and blurry images with noise. The model demonstrates high robustness to different conditions and is able to detect the edges of the bubbles and classify them accurately. Moreover, the model achieves an average intersection over union of 85%, indicating a high level of accuracy in predicting the masks of the bubbles. The method enables accurate recognition and tracking of individual bubble dynamics, capturing their coalescence, oscillation, and collisions to estimate local parameters by proving the bubble masks. This allows for a comprehensive understanding of their spatial-temporal behaviour, including the estimation of local Reynolds numbers.
|
|
| 8. |
- Soibam, Jerol
(författare)
-
Data-Driven Techniques for Fluid Mechanics and Heat Transfer
- 2022
-
Licentiatavhandling (övrigt vetenskapligt/konstnärligt)abstract
- One of the main challenges in fluid mechanics and heat transfer is the need for detailed studies andfast computational speed to monitor and optimise a system. These fluid/heat flows comprise time-dependent velocity, multi-scale, pressure, and energy fluctuations. Although there has been major advancements in computational power and technology, modelling detailed physical problems is currently falling short. The fluid mechanics and heat transfer domains are rapidly advancing, driven by unprecedented volumes of data from experiments, field measurements, and large-scale simulations at multiple spatio temporal scales. Such an increase in the volume of data unlocks the possibility of using techniques like machine learning. These machine learning algorithms offer a wealth of techniques to extract information from data that can be translated into knowledge about the underlying physics. Moreover, machine learning algorithms can augment domain knowledge and automate tasks related to flow control and optimisation. A significant milestone in the area of machine learning is the rise of deep learning, which is a powerful tool which can handle large data sets describing complex nonlinear dynamics that are commonly encountered in heat transfer and fluidflows.Therefore, this thesis aims to investigate data obtained from numerical simulations with deep learning techniques to reproduce the underlying physics present in data and considerably speed up the process. In this study, subcooled boiling transfer data has been used to train the deep neural network model then the trained model is validated using a validation dataset. The performance of the model is further evaluated using a set of interpolation and extrapolation datasets for different operating conditions outside the training and validation data. Furthermore, to highlight the robustness and reliability of the deep learning model, uncertainty quantification techniques such as Monte Carlo dropout and Deep Ensemble are implemented.This study demonstrates how a data-driven model can be used for subcooled boiling heat transfer and highlights why uncertainty quantification is important for such a model. The analysis and discussion in this thesis serve as the basis for further extending the potential use of data-driven methods for system optimisation, control and monitoring, diagnostic, and industrial applications.
|
|
| 9. |
- Soibam, Jerol, et al.
(författare)
-
Derivation and Uncertainty Quantification of a Data-Driven Subcooled Boiling Model
- 2020
-
Ingår i: Energies. - : MDPI. - 1996-1073. ; 13:22
-
Tidskriftsartikel (refereegranskat)abstract
- Subcooled flow boiling occurs in many industrial applications where enormous heat transfer is needed. Boiling is a complex physical process that involves phase change, two-phase flow, and interactions between heated surfaces and fluids. In general, boiling heat transfer is usually predicted by empirical or semiempirical models, which are horizontal to uncertainty. In this work, a data-driven method based on artificial neural networks has been implemented to study the heat transfer behavior of a subcooled boiling model. The proposed method considers the near local flow behavior to predict wall temperature and void fraction of a subcooled minichannel. The input of the network consists of pressure gradients, momentum convection, energy convection, turbulent viscosity, liquid and gas velocities, and surface information. The outputs of the models are based on the quantities of interest in a boiling system wall temperature and void fraction. To train the network, high-fidelity simulations based on the Eulerian two-fluid approach are carried out for varying heat flux and inlet velocity in the minichannel. Two classes of the deep learning model have been investigated for this work. The first one focuses on predicting the deterministic value of the quantities of interest. The second one focuses on predicting the uncertainty present in the deep learning model while estimating the quantities of interest. Deep ensemble and Monte Carlo Dropout methods are close representatives of maximum likelihood and Bayesian inference approach respectively, and they are used to derive the uncertainty present in the model. The results of this study prove that the models used here are capable of predicting the quantities of interest accurately and are capable of estimating the uncertainty present. The models are capable of accurately reproducing the physics on unseen data and show the degree of uncertainty when there is a shift of physics in the boiling regime.
|
|
| 10. |
- Soibam, Jerol, et al.
(författare)
-
Inverse flow prediction using ensemble PINNs and uncertainty quantification
- 2024
-
Ingår i: International Journal of Heat and Mass Transfer. - 0017-9310 .- 1879-2189. ; 226
-
Tidskriftsartikel (refereegranskat)abstract
- The thermal boundary conditions in a numerical simulation for heat transfer are often imprecise. This leads to poorly defined boundary conditions for the energy equation. The lack of accurate thermal boundary conditions in real-world applications makes it impossible to effectively solve the problem, regardless of the advancement of conventional numerical methods. This study utilises a physics-informed neural network to tackle ill-posed problems for unknown thermal boundaries with limited sensor data. The network approximates velocity and temperature fields while complying with the Navier-Stokes and energy equations, thereby revealing unknown thermal boundaries and reconstructing the flow field around a square cylinder. The method relies on optimal sensor placement determined by the QR pivoting technique, which ensures the effective capture of the dynamics, leading to enhanced model accuracy. In an effort to increase the robustness and generalisability, an ensemble physics-informed neural network is implemented. This approach mitigates the risks of overfitting and underfitting while providing a measure of model confidence. As a result, the ensemble model can identify regions of reliable prediction and potential inaccuracies. Therefore, broadening its applicability in tackling complex heat transfer problems with unknown boundary conditions.
|
|
