Please enable JavaScript.
Coggle requires JavaScript to display documents.
ML paper - Coggle Diagram
ML paper
Introduction
Background
Literature review
Main features of ST
Working principles
Torque production
Positive on ADV
Negative on RET
Performance augmentation
Common methods
Blade shape
AR
OR
Blade number
Minimizing negative torque
Fundamentals
Approaches
Externally
Guide vanes
Shielding plates
Deflectors
Internally
Static venting
Dynamic venting
Recognizing gaps
Static venting problems
Limited works
Maintaining omnidirectionally
Technical aspects
Limitations of CFD
Reducing computational time/cost
Physical model
New flap design
Removing flap gap
Realistic models
Mesh generation
Overset method
Improving mesh quality
Demand for optimization
Purpose
Problem statement
ML application
DOE
Scaling down size of data
Sampling data
ANN
Training
Testing
Prediction
GA
Combination of ML-GA
Reduce energy cost
Smaller flap size
Limited operational period
Enhance torque production
Reduce -torque on Ret
Preserve +pressure on Adv
Objectives
Dynamic venting approach
Improved flap models
Taguchi and ANOVA
Identifying Objective Function
Order contribution factor
Create database
Develop ANN predictive model
Apply GA for multi-objective optimization
Motivation
Broad vision
RE resources
Global warming issues
WT Specifics
WT categories
HAWT
VAWT
Lift-driven
Drag-driven
Advantages/Disadvantages
Significance of Savonius
Advantages
Omnidirectionality
Simplicity
Cost-effective
Problems
Low efficiency