Machine Learning: New Ideas and Tools in Environmental Science and Engineering

Zhong, Shifa; Zhang, Kai; Bagheri, Majid; Burken, Joel G.; Gu, April Z.; Li, Baikun; Ma, Xingmao; Marrone, Babetta L.; Ren, Zhiyong Jason; Schrier, Joshua; Shi, Wei; Tan, Haoyue; Wang, Tianbao; Wang, Xu; Wong, Bryan M.; Xiao, Xusheng; Yu, Xiong; Zhu, Jun‐Jie; Zhang, Huichun

doi:10.1021/acs.est.1c01339

Cited by 248 publications

(311 citation statements)

References 142 publications

Supporting

Mentioning

311

Contrasting

Order By: Relevance

“…The data collected were divided randomly into training and testing data sets in the algorithms for modelling as seen in Figure 18. It must however be highlighted that in recent times, there are more advanced machine learning techniques which tends to give superior outcomes and some of these algorithms may become primary point of focus for future work [37,38].…”

Section: Data Preparationmentioning

confidence: 99%

Effect of Bipolar Plate Material on Proton Exchange Membrane Fuel Cell Performance

et al. 2022

View full text Add to dashboard Cite

Commercialization of proton exchange membrane fuel cells can only materials provided its performance is closely related to existing technologies useful in commercial application. Other critical parameters like the utilization of cheaper materials should be taken into account during the manufacturing of the cell. A key component in the cell that has direct correlation to the cell performance is the flow plate. The weight coupled with cost of the cell revolves around the flow plate used in the manufacturing of the cell. This study explores materials ideal for the manufacturing of fuel cells in order to improve the overall cell performance. The investigation highlights the critical impact of varying materials used in the manufacturing of flow plates for PEM fuel cells. Stainless steel (SS), aluminium (Al) and copper (Cu) were the materials considered. The flow plate designs considered were serpentine and open pore cellular foam channel. Machine learning using python for the validation of the results with Linear regression, Ridge regression and Polynomial regression algorithm was carried out. The performance of both flow field channels was compared using different bipolar plate materials. The results show that metal foam flow channels overall performance was better than serpentine flow channels with all the various bipolar plate material used and Al material outperformed Cu and SS material. There is a direct correlation in terms of the outcome of the study and literature based on the data generated experimentally. It can however be concluded that molecules of hydrogen are stable on aluminium plates compared to copper and stainless steel.

show abstract

Section: Data Preparationmentioning

confidence: 99%

Effect of Bipolar Plate Material on Proton Exchange Membrane Fuel Cell Performance

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Existing research used skeleton joints as nodes to detect falls [121,326]. In addition, general human activity recognition (HAR) with wearable/smartphone sensors by using GNN and activity graphs [192,262] Traditional machine learning methods addressed many complex ESE problems [328], for example, like prediction of particulate matter(PM 2.5), water resource availability, and endocrine disrupting chemicals (EDCs), etc. The recent advancements in GNN better address these existing problems in the ESE field.…”

Section: Health Inference and Informaticsmentioning

confidence: 99%

Graph Neural Networks in IoT: A Survey

Dong¹,

Tang²,

Wang³

et al. 2022

Preprint

View full text Add to dashboard Cite

The Internet of Things (IoT) boom has revolutionized almost every corner of people's daily lives: healthcare, home, transportation, manufacturing, supply chain, and so on. With the recent development of sensor and communication technologies, IoT devices including smart wearables, cameras, smartwatches, and autonomous vehicles can accurately measure and perceive their surrounding environment. Continuous sensing generates massive amounts of data and presents challenges for machine learning. Deep learning models (e.g., convolution neural networks and recurrent neural networks) have been extensively employed in solving IoT tasks by learning patterns from multi-modal sensory data. Graph Neural Networks (GNNs), an emerging and fast-growing family of neural network models, can capture complex interactions within sensor topology and have been demonstrated to achieve state-of-the-art results in numerous IoT learning tasks. In this survey, we present a comprehensive review of recent advances in the application of GNNs to the IoT field, including a deep dive analysis of GNN design in various IoT sensing environments, an overarching list of public data and source code from the collected publications, and future research directions. To keep track of newly published works, we collect representative papers and their open-source implementations and create a Github repository at GNN4IoT. CCS Concepts: • Human-centered computing → Ubiquitous and mobile computing systems and tools; • Computing methodologies → Neural networks.

show abstract

“…Tailoring the mechanical properties can be achieved either by combining monomers with diverse chemistries to form either copolymers or blends [22][23][24][25] or by adding plasticizers to the polymer matrix [26,27]. To accelerate the rational design of PHA-based polymers, the chemical design space has to be efficiently explored and the structure-property relationships established to identify new polymers with enhanced properties [28].…”

Section: Introductionmentioning

confidence: 99%

Predicting the Mechanical Response of Polyhydroxyalkanoate Biopolymers Using Molecular Dynamics Simulations

et al. 2022

Self Cite

View full text Add to dashboard Cite

Polyhydroxyalkanoates (PHAs) have emerged as a promising class of biosynthesizable, biocompatible, and biodegradable polymers to replace petroleum-based plastics for addressing the global plastic pollution problem. Although PHAs offer a wide range of chemical diversity, the structure–property relationships in this class of polymers remain poorly established. In particular, the available experimental data on the mechanical properties is scarce. In this contribution, we have used molecular dynamics simulations employing a recently developed forcefield to predict chemical trends in mechanical properties of PHAs. Specifically, we make predictions for Young’s modulus, and yield stress for a wide range of PHAs that exhibit varying lengths of backbone and side chains as well as different side chain functional groups. Deformation simulations were performed at six different strain rates and six different temperatures to elucidate their influence on the mechanical properties. Our results indicate that Young’s modulus and yield stress decrease systematically with increase in the number of carbon atoms in the side chain as well as in the polymer backbone. In addition, we find that the mechanical properties were strongly correlated with the chemical nature of the functional group. The functional groups that enhance the interchain interactions lead to an enhancement in both the Young’s modulus and yield stress. Finally, we applied the developed methodology to study composition-dependence of the mechanical properties for a selected set of binary and ternary copolymers. Overall, our work not only provides insights into rational design rules for tailoring mechanical properties in PHAs, but also opens up avenues for future high throughput atomistic simulation studies geared towards identifying functional PHA polymer candidates for targeted applications.

show abstract

Machine Learning: New Ideas and Tools in Environmental Science and Engineering

Cited by 248 publications

References 142 publications

Effect of Bipolar Plate Material on Proton Exchange Membrane Fuel Cell Performance

Effect of Bipolar Plate Material on Proton Exchange Membrane Fuel Cell Performance

Graph Neural Networks in IoT: A Survey

Predicting the Mechanical Response of Polyhydroxyalkanoate Biopolymers Using Molecular Dynamics Simulations

Contact Info

Product

Resources

About