Improving Machine Learning Model Performance in Predicting the Indicated Mean Effective Pressure of a Natural Gas Engine

2021 ◽  
Author(s):  
Jinlong Liu ◽  
Christopher Ulishney ◽  
Cosmin Dumitrescu
Keyword(s):  
Machine Learning ◽  
Natural Gas ◽  
Model Performance ◽  
Learning Model ◽  
Effective Pressure ◽  
Gas Engine ◽  
Natural Gas Engine ◽  
Indicated Mean Effective Pressure ◽  
Machine Learning Model

Improving Machine Learning Model Performance in Predicting the Indicated Mean Effective Pressure of a Natural Gas Engine

2020 ◽  
Author(s):  
Jinlong Liu ◽  
Christopher Ulishney ◽  
Cosmin E. Dumitrescu
Keyword(s):  
Machine Learning ◽  
Natural Gas ◽  
Decision Trees ◽  
Engine Speed ◽  
Model Performance ◽  
Learning Model ◽  
Operating Conditions ◽  
Effective Pressure ◽  
Indicated Mean Effective Pressure ◽  
Machine Learning Model

Abstract Converting existing compression ignition engines to spark ignition approach is a promising approach to increase the application of natural gas in the heavy-duty transportation sector. However, the diesel-like environment dramatically affects the engine performance and emissions. As a result, experimental tests are needed to investigate the characteristics of such converted engines. A machine learning model based on bagged decision trees algorithm was established in this study to reduce the experimental cost and identify the operating conditions of special interest for analysis. Preliminary engine tests that changed spark timing, mixture equivalence ratio, and engine speed (three key engine operation variables) but maintained intake and boundary conditions were applied as model input to train such a correlative model. The model output was the indicated mean effective pressure, which is an engine parameter generally used to assist in locating high engine efficiency regions at constant engine speed and fuel/air ratio. After training, the correlative model can provide acceptable prediction performance except few outliers. Subsequently, boosting ensemble learning approach was applied in this study to help improve the model performance. Furthermore, the results showed that the boosted decision trees algorithm better described the combustion process inside the cylinder, as least for the operating conditions investigated in this study.

Response surface method optimization of a natural gas engine with dedicated exhaust gas recirculation

2021 ◽  
pp. 146808742199652
Author(s):  
Chris A Van Roekel ◽  
David T Montgomery ◽  
Jaswinder Singh ◽  
Daniel B Olsen
Keyword(s):  
Natural Gas ◽  
Operating Conditions ◽  
Exhaust Gas ◽  
Effective Pressure ◽  
Gas Engine ◽  
Natural Gas Engine ◽  
Indicated Mean Effective Pressure ◽  
Surface Method ◽  
Rsm Optimization ◽  
Gas Recirculation

Stoichiometric industrial natural gas engines rely on robust design to achieve consumer driven up-time requirements. Key to this design are exhaust components that are able to withstand high combustion temperatures found in this type of natural gas engine. The issue of exhaust component durability can be addressed by making improvements to materials and coatings or decreasing combustion temperatures. Among natural gas engine technologies shown to reduce combustion temperature, dedicated exhaust gas recirculation (EGR) has limited published research. However, due to the high nominal EGR rate it may be a technology useful for decreasing combustion temperature. In previous work by the author, dedicated EGR was implemented on a Caterpillar G3304 stoichiometric natural gas engine. Examination of combustion statistics showed that, in comparison to a conventional stoichiometric natural gas engine, operating with dedicated EGR requires adjustments to the combustion recipe to achieve acceptable engine operation. This work focuses on modifications to the combustion recipe necessary to improve combustion statistics such as coefficient of variance of indicated mean effective pressure (COV of IMEP), cylinder-cylinder indicated mean effective pressure (IMEP), location of 50% mass fraction burned, and 10%–90% mass fraction burn duration. Several engine operating variables were identified to affect these combustion statistics. A response surface method (RSM) optimization was chosen to find engine operating conditions that would result in improved combustion statistics. A third order factorial RSM optimization was sufficient for finding optimized operating conditions at 3.4 bar brake mean effective pressure (BMEP). The results showed that in an engine with a low turbulence combustion chamber, such as a G3304, optimized combustion statistics resulted from a dedicated cylinder lambda of 0.936, spark timing of 45° before top dead center (°bTDC), spark duration of 365 µs, and intake manifold temperature of 62°C. These operating conditions reduced dedicated cylinder COV of IMEP by 10% (absolute) and the difference between average stoichiometric cylinder and dedicated cylinder IMEP to 0.19 bar.

scholarly journals Assessing the International Transferability of a Machine Learning Model for Detecting Medication Error in the General Internal Medicine Clinic: Multicenter Preliminary Validation Study

10.2196/23454 ◽  
2021 ◽  
Vol 9 (1) ◽  
pp. e23454
Author(s):  
Yen Po Harvey Chin ◽  
Wenyu Song ◽  
Chia En Lien ◽  
Chang Ho Yoon ◽  
Wei-Chen Wang ◽  
...  
Keyword(s):  
Machine Learning ◽  
Medication Error ◽  
Model Performance ◽  
Learning Model ◽  
Study Cohort ◽  
Learning Approach ◽  
Hospital Data ◽  
General Internal ◽  
Machine Learning Model ◽  
Outpatient Prescriptions

Background Although most current medication error prevention systems are rule-based, these systems may result in alert fatigue because of poor accuracy. Previously, we had developed a machine learning (ML) model based on Taiwan’s local databases (TLD) to address this issue. However, the international transferability of this model is unclear. Objective This study examines the international transferability of a machine learning model for detecting medication errors and whether the federated learning approach could further improve the accuracy of the model. Methods The study cohort included 667,572 outpatient prescriptions from 2 large US academic medical centers. Our ML model was applied to build the original model (O model), the local model (L model), and the hybrid model (H model). The O model was built using the data of 1.34 billion outpatient prescriptions from TLD. A validation set with 8.98% (60,000/667,572) of the prescriptions was first randomly sampled, and the remaining 91.02% (607,572/667,572) of the prescriptions served as the local training set for the L model. With a federated learning approach, the H model used the association values with a higher frequency of co-occurrence among the O and L models. A testing set with 600 prescriptions was classified as substantiated and unsubstantiated by 2 independent physician reviewers and was then used to assess model performance. Results The interrater agreement was significant in terms of classifying prescriptions as substantiated and unsubstantiated (κ=0.91; 95% CI 0.88 to 0.95). With thresholds ranging from 0.5 to 1.5, the alert accuracy ranged from 75%-78% for the O model, 76%-78% for the L model, and 79%-85% for the H model. Conclusions Our ML model has good international transferability among US hospital data. Using the federated learning approach with local hospital data could further improve the accuracy of the model.

scholarly journals The importance of round-robin validation when assessing machine-learning-based vertical extrapolation of wind speeds

2020 ◽  
Author(s):  
Nicola Bodini ◽  
Mike Optis
Keyword(s):  
Machine Learning ◽  
Random Forest ◽  
Power Law ◽  
Wind Farm ◽  
Sonic Anemometer ◽  
Model Performance ◽  
Learning Model ◽  
Round Robin ◽  
Wind Speeds ◽  
Machine Learning Model

Abstract. The extrapolation of wind speeds measured at a meteorological mast to wind turbine hub heights is a key component in a bankable wind farm energy assessment and a significant source of uncertainty. Industry-standard methods for extrapolation include the power law and logarithmic profile. The emergence of machine-learning applications in wind energy has led to several studies demonstrating substantial improvements in vertical extrapolation accuracy in machine-learning methods over these conventional power law and logarithmic profile methods. In all cases, these studies assess relative model performance at a measurement site where, critically, the machine-learning algorithm requires knowledge of the hub-height wind speeds in order to train the model. This prior knowledge provides fundamental advantages to the site-specific machine-learning model over the power law and log profile, which, by contrast, are not highly tuned to hub-height measurements but rather can generalize to any site. Furthermore, there is no practical benefit in applying a machine-learning model at a site where hub-height winds are known; rather, its performance at nearby locations (i.e., across a wind farm site) without hub-height measurements is of most practical interest. To more fairly and practically compare machine-learning-based extrapolation to standard approaches, we implemented a round-robin extrapolation model comparison, in which a random forest machine-learning model is trained and evaluated at different sites and then compared against the power law and logarithmic profile. We consider 20 months of lidar and sonic anemometer data collected at four sites between 50–100 kilometers apart in the central United States. We find that the random forest outperforms the standard extrapolation approaches, especially when incorporating surface measurements as inputs to include the influence of atmospheric stability. When compared at a single site (the traditional comparison approach), the machine-learning improvement in mean absolute error was 28 % and 23 % over the power law and logarithmic profile, respectively. Using the round-robin approach proposed here, this improvement drops to 19 % and 14 %, respectively. These latter values better represent practical model performance, and we conclude that round-robin validation should be the standard for machine-learning-based, wind-speed extrapolation methods.

Random Forest Machine Learning Model for Predicting Combustion Feedback Information of a Natural Gas Spark Ignition Engine

2020 ◽  
Vol 143 (1) ◽  
Author(s):  
Jinlong Liu ◽  
Christopher Ulishney ◽  
Cosmin Emil Dumitrescu
Keyword(s):  
Machine Learning ◽  
Random Forest ◽  
Natural Gas ◽  
Engine Performance ◽  
Cost Effective ◽  
Learning Model ◽  
Operating Conditions ◽  
Control Variables ◽  
Feedback Information ◽  
Machine Learning Model

Abstract Engine calibration requires detailed feedback information that can reflect the combustion process as the optimized objective. Indicated mean effective pressure (IMEP) is such an indicator describing an engine’s capacity to do work under different combinations of control variables. In this context, it is of interest to find cost-effective solutions that will reduce the number of experimental tests. This paper proposes a random forest machine learning model as a cost-effective tool for optimizing engine performance. Specifically, the model estimated IMEP for a natural gas spark ignited engine obtained from a converted diesel engine. The goal was to develop an economical and robust tool that can help reduce the large number of experiments usually required throughout the design and development of internal combustion engines. The data used for building such correlative model came from engine experiments that varied the spark advance, fuel-air ratio, and engine speed. The inlet conditions and the coolant/oil temperature were maintained constant. As a result, the model inputs were the key engine operation variables that affect engine performance. The trained model was shown to be able to predict the combustion-related feedback information with good accuracy (R2 ≈ 0.9 and MSE ≈ 0). In addition, the model accurately reproduced the effect of control variables on IMEP, which would help narrow the choice of operating conditions for future designs of experiment. Overall, the machine learning approach presented here can provide new chances for cost-efficient engine analysis and diagnostics work.

scholarly journals Assessing the International Transferability of a Machine Learning Model for Detecting Medication Error in the General Internal Medicine Clinic: Multicenter Preliminary Validation Study (Preprint)

2020 ◽  
Author(s):  
Yen Po Harvey Chin ◽  
Wenyu Song ◽  
Chia En Lien ◽  
Chang Ho Yoon ◽  
Wei-Chen Wang ◽  
...  
Keyword(s):  
Machine Learning ◽  
Medication Error ◽  
Model Performance ◽  
Learning Model ◽  
Study Cohort ◽  
Learning Approach ◽  
Hospital Data ◽  
General Internal ◽  
Machine Learning Model ◽  
Outpatient Prescriptions

BACKGROUND Although most current medication error prevention systems are rule-based, these systems may result in alert fatigue because of poor accuracy. Previously, we had developed a machine learning (ML) model based on Taiwan’s local databases (TLD) to address this issue. However, the international transferability of this model is unclear. OBJECTIVE This study examines the international transferability of a machine learning model for detecting medication errors and whether the federated learning approach could further improve the accuracy of the model. METHODS The study cohort included 667,572 outpatient prescriptions from 2 large US academic medical centers. Our ML model was applied to build the original model (O model), the local model (L model), and the hybrid model (H model). The O model was built using the data of 1.34 billion outpatient prescriptions from TLD. A validation set with 8.98% (60,000/667,572) of the prescriptions was first randomly sampled, and the remaining 91.02% (607,572/667,572) of the prescriptions served as the local training set for the L model. With a federated learning approach, the H model used the association values with a higher frequency of co-occurrence among the O and L models. A testing set with 600 prescriptions was classified as <i>substantiated</i> and <i>unsubstantiated</i> by 2 independent physician reviewers and was then used to assess model performance. RESULTS The interrater agreement was significant in terms of classifying prescriptions as <i>substantiated</i> and <i>unsubstantiated</i> (κ=0.91; 95% CI 0.88 to 0.95). With thresholds ranging from 0.5 to 1.5, the alert accuracy ranged from 75%-78% for the O model, 76%-78% for the L model, and 79%-85% for the H model. CONCLUSIONS Our ML model has good international transferability among US hospital data. Using the federated learning approach with local hospital data could further improve the accuracy of the model.

scholarly journals New methods to improve the vertical extrapolation of near-surface offshore wind speeds

2021 ◽  
Author(s):  
Mike Optis ◽  
Nicola Bodini ◽  
Mithu Debnath ◽  
Paula Doubrawa
Keyword(s):  
Machine Learning ◽  
Wind Speed ◽  
Model Performance ◽  
Learning Model ◽  
Offshore Wind ◽  
Offshore Wind Turbine ◽  
Near Surface ◽  
Wind Speeds ◽  
Machine Learning Model ◽  
Surface Measurements

Abstract. Accurate characterization of the offshore wind resource has been hindered by a sparsity of wind speed observations that span offshore wind turbine rotor-swept heights. Although public availability of floating lidar data is increasing, most offshore wind speed observations continue to come from buoy-based and satellite-based near-surface measurements. The aim of this study is to develop and validate novel vertical extrapolation methods that can accurately estimate wind speed time series across rotor-swept heights using these near-surface measurements. We contrast the conventional logarithmic profile against three novel approaches: a logarithmic profile with a long-term stability correction, a single-column model, and a machine-learning model. These models are developed and validated using 1 year of observations from two floating lidars deployed in U.S. Atlantic offshore wind energy areas. We find that the machine-learning model significantly outperforms all other models across all stability regimes, seasons, and times of day. Machine-learning model performance is considerably improved by including the air-sea temperature difference, which provides some accounting for offshore atmospheric stability. Finally, we find no degradation in machine-learning model performance when tested 83 km from its training location, suggesting promising future applications in extrapolating 10-m wind speeds from spatially resolved satellite-based wind atlases.

scholarly journals New methods to improve the vertical extrapolation of near-surface offshore wind speeds

2021 ◽  
Vol 6 (3) ◽  
pp. 935-948
Author(s):  
Mike Optis ◽  
Nicola Bodini ◽  
Mithu Debnath ◽  
Paula Doubrawa
Keyword(s):  
Machine Learning ◽  
Wind Speed ◽  
Model Performance ◽  
Learning Model ◽  
Offshore Wind ◽  
Offshore Wind Turbine ◽  
Near Surface ◽  
Wind Speeds ◽  
Machine Learning Model ◽  
Surface Measurements

Abstract. Accurate characterization of the offshore wind resource has been hindered by a sparsity of wind speed observations that span offshore wind turbine rotor-swept heights. Although public availability of floating lidar data is increasing, most offshore wind speed observations continue to come from buoy-based and satellite-based near-surface measurements. The aim of this study is to develop and validate novel vertical extrapolation methods that can accurately estimate wind speed time series across rotor-swept heights using these near-surface measurements. We contrast the conventional logarithmic profile against three novel approaches: a logarithmic profile with a long-term stability correction, a single-column model, and a machine-learning model. These models are developed and validated using 1 year of observations from two floating lidars deployed in US Atlantic offshore wind energy areas. We find that the machine-learning model significantly outperforms all other models across all stability regimes, seasons, and times of day. Machine-learning model performance is considerably improved by including the air–sea temperature difference, which provides some accounting for offshore atmospheric stability. Finally, we find no degradation in machine-learning model performance when tested 83 km from its training location, suggesting promising future applications in extrapolating 10 m wind speeds from spatially resolved satellite-based wind atlases.

Sign in / Sign up

Export Citation Format

Share Document