scholarly journals Artificial neural networks for inverse design of resonant nanophotonic components with oscillatory loss landscapes

Nanophotonics ◽  
2020 ◽  
Vol 10 (1) ◽  
pp. 385-392
Author(s):  
Joeri Lenaerts ◽  
Hannah Pinson ◽  
Vincent Ginis
Keyword(s):  
Machine Learning ◽  
Optimization Procedure ◽  
Linear Mapping ◽  
Inverse Design ◽  
Design Parameters ◽  
Second Step ◽  
Input Parameters ◽  
Novel Variant ◽  
Resonant Systems ◽  
The Fourier Transform

AbstractMachine learning offers the potential to revolutionize the inverse design of complex nanophotonic components. Here, we propose a novel variant of this formalism specifically suited for the design of resonant nanophotonic components. Typically, the first step of an inverse design process based on machine learning is training a neural network to approximate the non-linear mapping from a set of input parameters to a given optical system’s features. The second step starts from the desired features, e.g. a transmission spectrum, and propagates back through the trained network to find the optimal input parameters. For resonant systems, this second step corresponds to a gradient descent in a highly oscillatory loss landscape. As a result, the algorithm often converges into a local minimum. We significantly improve this method’s efficiency by adding the Fourier transform of the desired spectrum to the optimization procedure. We demonstrate our method by retrieving the optimal design parameters for desired transmission and reflection spectra of Fabry–Pérot resonators and Bragg reflectors, two canonical optical components whose functionality is based on wave interference. Our results can be extended to the optimization of more complex nanophotonic components interacting with structured incident fields.

scholarly journals Simulator Acceleration and Inverse Design of Fin Field-Effect Transistors Using Machine Learning

2021 ◽  
Author(s):  
Insoo Kim ◽  
So Jeong Park ◽  
Changwook Jeong ◽  
Munbo Shim ◽  
Dae Sin Kim ◽  
...  
Keyword(s):  
Machine Learning ◽  
Field Effect ◽  
Field Effect Transistors ◽  
Semiconductor Industry ◽  
Electronic Devices ◽  
Efficient Solutions ◽  
Inverse Design ◽  
Design Parameters ◽  
Learning Approaches ◽  
Design Problems

Abstract The simulation and design of electronic devices such as transistors is vital for the semiconductor industry. Conventionally, a device is intuitively designed and simulated using model equations, which is a time-consuming and expensive process. However, recent machine learning approaches provide an unprecedented opportunity to improve these tasks by training the underlying relationships between the device design and the specifications derived from the extensively accumulated simulation data. This study implements various machine learning approaches for the simulation acceleration and inverse-design problems of fin field-effect transistors. In comparison to traditional simulators, the proposed neural network model demonstrated almost equivalent results (R2 = 0.99) and was more than 122,000 times faster in simulation. Moreover, the proposed inverse-design model successfully generated design parameters that satisfied the desired target specifications with high accuracies (R2 = 0.96). Overall, the results demonstrated that the proposed machine learning models aided in achieving efficient solutions for the simulation and design problems pertaining to electronic devices. Thus, the proposed approach can be further extended to more complex devices and other vital processes in the semiconductor industry.

scholarly journals Ensemble Machine Learning Assisted Reservoir Characterization Using Field Production Data–An Offshore Field Case Study

Energies ◽  
2021 ◽  
Vol 14 (4) ◽  
pp. 1052
Author(s):  
Baozhong Wang ◽  
Jyotsna Sharma ◽  
Jianhua Chen ◽  
Patricia Persaud
Keyword(s):  
Machine Learning ◽  
Random Forest ◽  
Reservoir Characterization ◽  
Time Lapse ◽  
Production Data ◽  
Oil Saturation ◽  
Ensemble Machine Learning ◽  
Input Parameters ◽  
Saturation Profiles ◽  
Field Production

Estimation of fluid saturation is an important step in dynamic reservoir characterization. Machine learning techniques have been increasingly used in recent years for reservoir saturation prediction workflows. However, most of these studies require input parameters derived from cores, petrophysical logs, or seismic data, which may not always be readily available. Additionally, very few studies incorporate the production data, which is an important reflection of the dynamic reservoir properties and also typically the most frequently and reliably measured quantity throughout the life of a field. In this research, the random forest ensemble machine learning algorithm is implemented that uses the field-wide production and injection data (both measured at the surface) as the only input parameters to predict the time-lapse oil saturation profiles at well locations. The algorithm is optimized using feature selection based on feature importance score and Pearson correlation coefficient, in combination with geophysical domain-knowledge. The workflow is demonstrated using the actual field data from a structurally complex, heterogeneous, and heavily faulted offshore reservoir. The random forest model captures the trends from three and a half years of historical field production, injection, and simulated saturation data to predict future time-lapse oil saturation profiles at four deviated well locations with over 90% R-square, less than 6% Root Mean Square Error, and less than 7% Mean Absolute Percentage Error, in each case.

Using Artificial Intelligence to Estimate Nonlinear Resilient Modulus Parameters from Common Index Properties

2021 ◽  
pp. 036119812110237
Author(s):  
Laura Camarena
Keyword(s):  
Machine Learning ◽  
Material Properties ◽  
Resilient Modulus ◽  
Nonlinear Behavior ◽  
Machine Learning Techniques ◽  
Design Parameters ◽  
Nonlinear Parameters ◽  
State Highway ◽  
Learning Techniques ◽  
State Highway Agencies

The Mechanistic–Empirical Pavement Design Guide (MEPDG) considers a hierarchical approach to determine the input values necessary for most design parameters. Level 1 requires site-specific measurement of the material properties from laboratory testing, whereas other levels make use of equations developed from regression models to estimate the material properties. Resilient modulus is a mechanical property that characterizes the unbound and subgrade materials under loading that is essential for the mechanistic design of pavements. The MEPDG resilient modulus model makes use of a three-parameter constitutive model to characterize the nonlinear behavior of the geomaterials. As the resilient modulus tests are complex, expensive, and require lengthy preparation time, most state highway agencies are unlikely to implement them as routine daily applications. Therefore, it is imperative to make use of models to calculate these nonlinear parameters. Existing models to determine these parameters are frequently based on linear regression. With the development of machine learning techniques, it is feasible to develop simpler equations that can be used to estimate the nonlinear parameters more accurately. This study makes use of the Long-Term Pavement Performance database and machine learning techniques to improve the equations utilized to determine the nonlinear parameters crucial to estimate the resilient modulus of unbound base and subgrade materials.

scholarly journals Tackling Photonic Inverse Design with Machine Learning

2021 ◽  
pp. 2002923
Author(s):  
Zhaocheng Liu ◽  
Dayu Zhu ◽  
Lakshmi Raju ◽  
Wenshan Cai
Keyword(s):  
Machine Learning ◽  
Inverse Design

Optimization of Turbofan Propulsion Cycle Using a Genetic Algorithm

2010 ◽  
Author(s):  
Adel Ghenaiet
Keyword(s):  
Genetic Algorithm ◽  
Fuel Consumption ◽  
Pressure Ratio ◽  
Engine Performance ◽  
Optimization Procedure ◽  
Specific Fuel Consumption ◽  
High Mass ◽  
Design Parameters ◽  
Inlet Temperature ◽  
Simplifying Assumptions

This paper presents an evolutionary approach as the optimization framework to design for the optimal performance of a high-bypass unmixed turbofan to match with the power requirements of a commercial aircraft. The parametric analysis had the objective to highlight the effects of the principal design parameters on the propulsive performance in terms of specific fuel consumption and specific thrust. The design optimization procedure based on the genetic algorithm PIKAIA coupled to the developed engine performance analyzer (on-design and off-design) aimed at finding the propulsion cycle parameters minimizing the specific fuel consumption, while meeting the required thrusts in cruise and takeoff and the restrictions of temperatures limits, engine size and weight as well as pollutants emissions. This methodology does not use engine components’ maps and operates on simplifying assumptions which are satisfying the conceptual or early design stages. The predefined requirements and design constraints have resulted in an engine with high mass flow rate, bypass ratio and overall pressure ratio and a moderate turbine inlet temperature. In general, the optimized engine is fairly comparable with available engines of equivalent power range.

scholarly journals Deep machine learning of the spectral power distribution of the LED system with multiple degradation mechanisms

2020 ◽  
Vol 37 ◽  
pp. 172-183
Author(s):  
Cadmus C A Yuan ◽  
JiaJie Fan ◽  
XueJun Fan
Keyword(s):  
Machine Learning ◽  
Thermal Aging ◽  
Empirical Relationship ◽  
Machine Learning Techniques ◽  
Color Temperature ◽  
Design Parameters ◽  
Degradation Mechanisms ◽  
Theoretical Understanding ◽  
Color Coordinate ◽  
Led Luminaire

Abstract The performance and reliability of the light-emitting diode (LED) system significantly depend on the thermal–mechanical loading-enhanced multiple degradation mechanisms and their interactions. The complexity of the LED system restricts the theoretical understanding of the root causes of the luminous fluctuation or the establishment of the direct correlation between the thermal aging loading and the luminous outputs. This paper applies the deep machine learning techniques and develops a gated network with the two-step learning algorithm to build the empirical relationship between the design parameters and the thermal aging loading and the luminous output of LED products. The flexibility of the proposed method will be demonstrated by integrating it with different neural network architectures. The proposed gated network concept has been validated in both multiple LED chip packaging and LED luminaire under thermal aging loading. The validation of the luminous data of multiple LED chip packaging shows that the maximum differences of the correlated color temperature (CCT) and color coordinate are 2.6% and 1.0%, respectively. Moreover, the machine learning results of the LED luminaire exhibit that the differences of lumen depreciation, CCT and color coordinate are 1.6%, 1.9% and 1.1%, after 2160 h of thermal aging.

Prediction of Threshold Sand Rates from Acoustic Monitors Using Artificial Intelligence

2021 ◽  
Author(s):  
Ronald E. Vieira ◽  
Bohan Xu ◽  
Asad Nadeem ◽  
Ahmed Nadeem ◽  
Siamack A. Shirazi
Keyword(s):  
Artificial Intelligence ◽  
Machine Learning ◽  
Noise Level ◽  
Background Noise ◽  
Pipe Wall ◽  
Flow Patterns ◽  
Operating Conditions ◽  
Flow Conditions ◽  
Background Noise Level ◽  
Input Parameters

Abstract Solids production from oil and gas wells can cause excessive damage resulting in safety hazards and expensive repairs. To prevent the problems associated with sand influx, ultrasonic devices can be used to provide a warning when sand is being produced in pipelines. One of the most used methods for sand detection is utilizing commercially available acoustic sand monitors that clamp to the outside of pipe wall and measures the acoustic energy generated by sand grain impacts on the inner side of a pipe wall. Although the transducer used by acoustic monitors is especially sensitive to acoustic emissions due to particle impact, it also reacts to flow induced noise as well (background noise). The acoustic monitor output does not exceed the background noise level until a sufficient sand rate is entrained in the flow that causes a signal output that is higher than the background noise level. This sand rate is referred to as the threshold sand rate or TSR. A significant amount of data has been compiled over the years for TSR at the Tulsa University Sand Management Projects (TUSMP) for various flow conditions with stainless steel pipe material. However, to use this data to develop a model for different flow patterns, fluid properties, pipe, and sand sizes is challenging. The purpose of this work is to develop an artificial intelligence (AI) methodology using machine learning (ML) models to determine TSR for a broad range of operating conditions. More than 250 cases from previous literature as well as ongoing research have been used to train and test the ML models. The data utilized in this work has been generated mostly in a large-scale multiphase flow loop for sand sizes ranging from 25 to 300 μm varying sand concentrations and pipe diameters from 25.4 mm to 101.6 mm ID in vertical and horizontal directions downstream of elbows. The ML algorithms including elastic net, random forest, support vector machine and gradient boosting, are optimized using nested cross-validation and the model performance is evaluated by R-squared score. The machine learning models were used to predict TSR for various velocity combinations under different flow patterns with sand. The sensitivity to changes of input parameters on predicted TSR was also investigated. The method for TSR prediction based on ML algorithms trained on lab data is also validated on actual field conditions available in the literature. The AI method results reveal a good training performance and prediction for a variety of flow conditions and pipe sizes not tested before. This work provides a framework describing a novel methodology with an expanded database to utilize Artificial Intelligence to correlate the TSR with the most common production input parameters.

Prediction of CO2 Minimum Miscibility Pressure Using an Augmented Machine-Learning-Based Model

SPE Journal ◽  
2021 ◽  
pp. 1-13
Author(s):  
Utkarsh Sinha ◽  
Birol Dindoruk ◽  
Mohamed Soliman
Keyword(s):  
Machine Learning ◽  
Phase Behavior ◽  
Hybrid Method ◽  
Gas Injection ◽  
Supervised Machine Learning ◽  
Design Parameters ◽  
Support Vector ◽  
Hydrocarbon Gases ◽  
Minimum Miscibility Pressure ◽  
Analytical Correlation

Summary Minimum miscibility pressure (MMP) is one of the key design parameters for gas injection projects. It is a physical parameter that is a measure of local displacement efficiency while subject to some constraints due to its definition. Also, the MMP value is used to tune compositional models along with proper fluid description constrained with other available basic phase behavior data, such as bubble point pressure and volumetric properties. In general, carbon dioxide (CO2) and hydrocarbon gases are the most common gases used for (or screened for) gas injection processes, and because of recent focus, they are used to screen for the coupling of CO2-sequestration and CO2-enhanced oil recovery (EOR) projects. Because the CO2/oil phase behavior is quite different than the hydrocarbon gas/oil phase behavior, researchers developed specialized correlations for CO2 or CO2-rich streams. Therefore, there is a need for a tool with expanded range capabilities for the estimation of MMP for CO2 gas streams. The only known and widely accepted measurement technique for MMP that is coherent with its formal definition is the use of a slimtube apparatus. However, the use of slimtube restricts the amount of data available, even though there are other alternative techniques presented over the last three decades, which all have various limitations (Dindoruk et al. 2021). Due to some of the complexities highlighted in Dindoruk et al. (2021) and time and resource requirements, there have been a number of correlations developed in the literature using mostly classical regression techniques with relatively sparse data using various combinations of limited input data (Cronquist 1978; Lee 1979; Yellig and Metcalfe 1980; Alston et al. 1985; Glaso 1985; Jaubert et al. 1998; Emera and Sarma 2005; Yuan et al. 2005; Ahmadi et al. 2010; Ahmadi and Johns 2011). In this paper, we present two separate approaches for the calculation of the MMP of an oil for CO2 injection: analytical correlation in which the correlation coefficients were tuned using linear support vector machines (SVMs) (Press et al. 2007; MathWorks 2020; RDocumentation 2020b; Cortes and Vapnik 1995) and using a hybrid method (i.e., superlearner model), which consists of the combination of random forest (RF) regression (Breiman 2001) and the proposed analytical correlation. Both models take the compositional analysis of oils up to heptane plus fraction, molecular weight of oil, and the reservoir temperature as input parameters. Based on statistical and data analysis techniques in combination with the help of corresponding crossplots, we showed that the performance of the final proposed method (hybrid method) is superior to all the leading correlations (Cronquist 1978; Lee 1979; Yellig and Metcalfe 1980; Alston et al. 1985; Glaso 1985; Emera and Sarma 2005; Yuan et al. 2005) and supervised machine-learning (Metcalfe 1982) methods considered in the literature (Altman 1992; Chambers and Hastie 1992; Chapelle and Vapnik 2000; Breiman 2001; Press et al. 2007; MathWorks 2020). The proposed model works for the widest spectrum of MMPs from 1,000 to 4,900 psia, which covers the entire range of oils within the scope of CO2 EOR based on the widely used screening criteria (Taber et al. 1997a, 1997b).

Vibration-Based Early Crack Diagnosis With Machine Learning for Spur Gears

2020 ◽  
Author(s):  
Fatih Karpat ◽  
Ahmet Emir Dirik ◽  
Onur Can Kalay ◽  
Oğuz Doğan ◽  
Burak Korcuklu
Keyword(s):  
Machine Learning ◽  
Fault Diagnosis ◽  
High Speed ◽  
Crack Detection ◽  
Power Transmission ◽  
Calculated Data ◽  
Design Parameters ◽  
Tooth Stiffness ◽  
Vibration Responses ◽  
Root Crack

Abstract Gear mechanisms are one of the most significant components of the power transmission systems. Due to increasing emphasis on the high-speed, longer working life, high torques, etc. cracks may be observed on the gear surface. Recently, Machine Learning (ML) algorithms have started to be used frequently in fault diagnosis with developing technology. The aim of this study is to determine the gear root crack and its degree with vibration-based diagnostics approach using ML algorithms. To perform early crack detection, the single tooth stiffness and the mesh stiffness calculated via ANSYS for both healthy and faulty (25-50-75-100%) teeth. The calculated data transferred to the 6-DOF dynamic model of a one-stage gearbox, and vibration responses was collected. The data gathered for healthy and faulty cases were evaluated for the feature extraction with five statistical indicators. Besides, white Gaussian noise was added to the data obtained from the 6-DOF model, and it was aimed at early fault diagnosis and condition monitoring with ML algorithms. In this study, the gear root crack and its degree analyzed for both healthy and four different crack sizes (25%-50%-75%-100%) for the gear crack detection. Thereby, a method was presented for early fault diagnosis without the need for a big experimental dataset. The proposed vibration-based approach can eliminate the high test rig construction costs and can potentially be used for the evaluation of different working conditions and gear design parameters. Therefore, catastrophic failures can be prevented, and maintenance costs can be optimized by early crack detection.

Sign in / Sign up

Export Citation Format

Share Document