More Like this

1. Initial Development of a Physics-Aware Machine Learning Framework for Soot Mass Prediction in Gasoline Direct Injection Engines

   https://doi.org/10.4271/2023-24-0174  

   Jayaprakash, Bharat; Wilmer, Brady; Northrop, William F (August 2023, 16th International Conference on Engines & Vehicles)

   Calibration of automotive engines to ensure compliance with emission regulations is a critical phase in product development. Control of engine-out particulate emissions, which directly impact the environment and public health, is particularly important. Detailed physics-based models are typically used to gain a rich understanding of the complex physical phenomena that drive the soot particle formation in an engine cylinder. However, such models often fail to correctly represent the highly dynamic nature of the underlying mechanisms under transient combustion conditions. Moreover, most physics-based models were initially developed for diesel engine applications and their applicability to gasoline engines remains questionable due to differences in flame structure and fuel-wall interactions. Black-box models have been previously proposed to predict engine-out soot emissions, but their lack of physical interpretability is an unsolved drawback. To address these limitations, we present a physics-aware twin-model machine learning framework to predict and analyze engine-out soot mass from a gasoline direct injection (GDI) engine. The framework combines a physics-based model with a bagging-type ensemble learning model that both maintains high accuracy and allows physical interpretation of results without using computationally intensive high-fidelity models. This work shows why a one-model-fits-all approach fails in the case of predicting soot emissions due to clustered co-occurrences of operating conditions that cause non-compliant behavior. We compare the performance of the proposed framework with that of the standalone baseline model and a feed-forward deep neural network. Using WLTP data from a 2.0L naturally aspirated GDI engine, the proposed framework predicts engine-out soot mass with an improvement of 29% in the R² value and 21% in the root mean squared error from the baseline physics-based model, without compromising physical interpretability. These improvements are significant enough to warrant further framework development with additional engine datasets. 

   more » « less
2. Implementation of CVR in Distribution Networks by Optimal Coordination of BESS and PV Inverters Using Arithmetic Optimization Algorithm

   https://doi.org/10.1109/NAPS52732.2021.9654489  

   Ahanch, Mojtaba; McCann, Roy (November 2021, IEEE Green Technologies Conference (GreenTech))

   This article proposes a new framework for the substation demand reduction and power loss minimization in distribution networks by implementing conservation voltage reduction (CVR) strategy. The proposed framework coordinates Battery Energy Storage Systems (BESS), Smart PV inverters and voltage control devices -including OLTC and voltage regulators- so that the substation demand and network power loss are reduced while the service voltage range meets the IEEE 1547 standard (120-114 V). The suggested CVR strategy is applied to the IEEE 34-bus case study system consisting of two PV generations and BESS. The smart PV inverters are controlled based on the combined Volt/VArVolt/Watt (VVW) characteristics scheme. Also, BESS is charged and discharged with regard to the time and peaks have control modes, respectively. The Arithmetic Optimization Algorithm (AOA) is implemented in MATLAB scripts for solving the optimization problem. Power flow studies are carried out using OpenDSS software. Results reveal that the new framework can achieve higher substation demand reduction considering the concurrent control of PVs and BESS. 

   more » « less
3. Dataset: Spatiotemporal variability and origin of CO2 and CH4 tree stem fluxes in an upland forest

   https://doi.org/10.6084/m9.figshare.14885697.v2  

   Barba, Josep; Poyatos, Rafael; Capooci, Margaret; Vargas, Rodrigo (January 2021, figshare)

   The exchange of multiple greenhouse gases (i.e., CO₂and CH4</sub>) between tree stems and the atmosphere represents a knowledge gap in the global carbon cycle. Stem CO2</sub> and CH4</sub> fluxes vary across time and space and is unclear which are their individual or shared drivers. This dataset contains information of CO2</sub> and CH4</sub> fluxes at different stem heights combining manual (biweekly; n=678) and automated (hourly; n>38,000) measurements in a temperate upland forest.</div>This study was performed in an upland forested area at the St. Jones Reserve [39°5’20”N, 75°26’21”W], a component of the Delaware National Estuarine Research Reserve (DNERR).</div></div>The dominant vegetation species are bitternut hickory (Carya cordiformis</i>), eastern red cedar (Juniperus virginiana</i> L.), American holly (Ilex opaca</i> (Ashe)), sweet gum (Liquidambar styraciflua</i> L.) and black gum (Nyssa sylvatica</i> (Marshall)), with an overall tree density of 678 stems ha^-1 and mean diameter at breast height (DBH) of 25.7±13.9 cm (mean±sd). We studied bitternut hickory, which is one of the most important species in the study site, accounting for 24.9% of the total basal area.</div></div>For code </div></sup></sub> 

   more » « less
4. Reinforcement Learning Based Fast Charging of Electric Vehicle Battery Packs

   https://doi.org/10.4271/2023-01-1681  

   Abbasi, Mohammad Hossein; Arjmandzadeh_PhD, Ziba; Zhang, Jiangfeng; Xu, Bin; Krovi, Venkat N (October 2023, Society of Automotive Engineers International)

   Range anxiety and lack of adequate access to fast charging are proving to be important impediments to electric vehicle (EV) adoption. While many techniques to fast charging EV batteries (model-based & model-free) have been developed, they have focused on a single Lithium-ion cell. Extensions to battery packs are scarce, often considering simplified architectures (e.g., series-connected) for ease of modeling. Computational considerations have also restricted fast-charging simulations to small battery packs, e.g., four cells (for both series and parallel connected cells). Hence, in this paper, we pursue a model-free approach based on reinforcement learning (RL) to fast charge a large battery pack (comprising 444 cells). Each cell is characterized by an equivalent circuit model coupled with a second-order lumped thermal model to simulate the battery behavior. After training the underlying RL, the developed model will be straightforward to implement with low computational complexity. In detail, we utilize a Proximal Policy Optimization (PPO) deep RL as the training algorithm. The RL is trained in such a way that the capacity loss due to fast charging is minimized. The pack’s highest cell surface temperature is considered an RL state, along with the pack’s state of charge. Finally, in a detailed case study, the results are compared with the constant current-constant voltage (CC-CV) approach, and the outperformance of the RL-based approach is demonstrated. Our proposed PPO model charges the battery as fast as a CC-CV with a 5C constant stage while maintaining the temperature as low as a CC-CV with a 4C constant stage. 

   more » « less
5. Containerization Approach for High-Fidelity Terramechanics Simulations

   https://doi.org/10.4271/2023-01-0105  

   Jadhav, Sanskruti Deepak; Salvi, Ameya; Kosaraju, Krishna Chaitanya; Smereka, Jonathon; Brudnak, Mark; Krovi, Venkat N; Gorsich, David (April 2023, SAE Mobilus)

   Integrated modeling of vehicle, tire and terrain is a fundamental challenge to be addressed for off-road autonomous navigation. The complexities arise due to lack of tools and techniques to predict the continuously varying terrain and environmental conditions and the resultant non-linearities. The solution to this challenge can now be found in the plethora of data driven modeling and control techniques that have gained traction in the last decade. Data driven modeling and control techniques rely on the system’s repeated interaction with the environment to generate a lot of data and then use a function approximator to fit a model for the physical system with the data. Getting good quality and quantity of data may involve extensive experimentation with the physical system impacting developer’s resource. The process is computationally expensive, and the overhead time required is high.High-fidelity simulators coupled with cloud-based containers can help ease the challenge of data ‘quality’ and ‘quantity’. Project Chrono is a multi-physics simulation engine that provides high-fidelity simulation capabilities with emphasis on flow and terrain modeling. With a host of libraries and APIs for industry accepted tools like MATLAB, Simulink and TensorFlow, Project Chrono proves to be a powerful research bed for data-driven modeling and control development for off-road navigation. Containers are lightweight virtual machines that take away repetitive configurations by setting up a computational environment, including all necessary dependencies and libraries. Docker encapsulates an end-to-end platform solution for heavy computation challenges of deep learning applications and allows fast development and testing. The synergy between the high-fidelity simulator and the compute outsourcing capabilities of cloud-based containers proves to be extremely beneficial for continuous integration and continuous deployment (CI/CD) for data driven modeling and control tasks. In the following work, we containerize a high-fidelity simulator (Project Chrono) to develop and validate data driven modeling and control algorithms for off-road autonomous navigation. 

   more » « less