scholarly journals Sensitivity of Typhoon Simulation to Physics Parameterizations in the Global Model

Atmosphere ◽  
2017 ◽  
Vol 27 (1) ◽  
pp. 17-28 ◽  
Author(s):  
Ki-Byung Kim ◽  
Eun-Hee Lee ◽  
Kyung-Hee Seol
Keyword(s):  
Global Model ◽  
Physics Parameterizations

scholarly journals FedPSO: Federated Learning Using Particle Swarm Optimization to Reduce Communication Costs

Sensors ◽  
2021 ◽  
Vol 21 (2) ◽  
pp. 600
Author(s):  
Sunghwan Park ◽  
Yeryoung Suh ◽  
Jaewoo Lee
Keyword(s):  
Particle Swarm Optimization ◽  
Data Privacy ◽  
Particle Swarm ◽  
Global Model ◽  
Network Communication ◽  
Improve Performance ◽  
Communication Performance ◽  
Swarm Optimization ◽  
Model Update ◽  
Network Environments

Federated learning is a learning method that collects only learned models on a server to ensure data privacy. This method does not collect data on the server but instead proceeds with data directly from distributed clients. Because federated learning clients often have limited communication bandwidth, communication between servers and clients should be optimized to improve performance. Federated learning clients often use Wi-Fi and have to communicate in unstable network environments. However, as existing federated learning aggregation algorithms transmit and receive a large amount of weights, accuracy is significantly reduced in unstable network environments. In this study, we propose the algorithm using particle swarm optimization algorithm instead of FedAvg, which updates the global model by collecting weights of learned models that were mainly used in federated learning. The algorithm is named as federated particle swarm optimization (FedPSO), and we increase its robustness in unstable network environments by transmitting score values rather than large weights. Thus, we propose a FedPSO, a global model update algorithm with improved network communication performance, by changing the form of the data that clients transmit to servers. This study showed that applying FedPSO significantly reduced the amount of data used in network communication and improved the accuracy of the global model by an average of 9.47%. Moreover, it showed an improvement in loss of accuracy by approximately 4% in experiments on an unstable network.

Global model analysis of Ar inductively coupled plasma driven by a 150 kHz-band high-power pulse burst

2019 ◽  
Vol 58 (SA) ◽  
pp. SAAB06
Author(s):  
Yuma Saito ◽  
Kodai Shibata ◽  
Katsuyuki Takahashi ◽  
Seiji Mukaigawa ◽  
Koichi Takaki ◽  
...  
Keyword(s):  
High Power ◽  
Inductively Coupled Plasma ◽  
Model Analysis ◽  
Global Model ◽  
Pulse Burst ◽  
Inductively Coupled ◽  
Power Pulse ◽  
High Power Pulse

scholarly journals Modeling of the buckstay system of membrane-walls in watertube boiler construction

2014 ◽  
Vol 18 (suppl.1) ◽  
pp. 59-72 ◽  
Author(s):  
Hasan Nagiar ◽  
Tasko Maneski ◽  
Vesna Milosevic-Mitic ◽  
Branka Gacesa ◽  
Nina Andjelic
Keyword(s):  
Finite Element ◽  
Orthotropic Plate ◽  
Global Model ◽  
Local Model ◽  
Thermal Loads ◽  
Water Tube ◽  
Bending Load ◽  
Specific Geometry ◽  
Structural Parts ◽  
Help Model

Membrane walls are very important structural parts of water-tube boiler construction. Based on their specific geometry, one special type of finite element was defined to help model the global boiler construction. That is the element of reduced orthotropic plate with two thicknesses and two elasticity matrixes, for membrane and bending load separately. A global model of the boiler construction showed that the high value of stress is concentrated in plates of the buckstay system in boiler corners. Validation of the new finite element was done on the local model of the part of membrane wall and buckstay. A very precise model of tubes and flanges was compared to the model formed on the element of a reduced orthotropic plate. Pressure and thermal loads were discussed. Obtained results indicated that the defined finite element was quite favorable in the design and reconstruction of the boiler substructures such as a buckstay system.

Sign in / Sign up

Export Citation Format

Share Document