Abstract
Wind tunnel tests are crucial for studying wind turbine wake effects, but the Reynolds effect causes discrepancies in
power and thrust coefficients between scaled models and full-scale turbines, impacting wake simulation accuracy. This paper
proposes a bi-objective optimization design method of the scaled blade based on the thrust-power coefficient. Using a pattern
search algorithm, the method optimizes both coefficients by adjusting optimization objectives. Results show that chord length
distribution primarily affects the thrust coefficient when airfoil type is fixed, while twist angle significantly influences the power
coefficient. The chord length and twist angle are then considered as optimization variables to design a scaled blade. After
optimization, the thrust coefficient is consistent with the prototype turbine, while the power coefficient reaches 0.2, which could
be used in the 5MW wind turbine test. Furthermore, the effects of chord length and twist angle on the thrust and power
coefficients of the wind turbine blade were examined using a 5MW wind turbine as an example, respectively. In this
examination, numerical simulation is employed to verify the aerodynamic parameters of the optimized blade with thrust-power
coefficient, thereby demonstrating the reliability of bi-objective optimization method.
Key Words
bi-objective optimization; blade design; numerical simulation; thrust-power coefficient
Address
Senqin Zhang: 1)School of Civil Engineering, Chongqing University, Chongqing 400044, China
2)Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, China
Binbin Wang: PowerChina Sichuan electric power engineering Co., Ltd
Guoqing Huang: School of Civil Engineering, Chongqing University, Chongqing 400044, China
Bowen Yan: School of Civil Engineering, Chongqing University, Chongqing 400044, China
Guowei Qian: School of Ocean Engineering and Technology, Sun Yat-sen University, Zhuhai 519082, China
Ke Li: School of Civil Engineering, Chongqing University, Chongqing 400044, China
Abstract
As the amount of airfoil optimization data and the complexity of models increase, the optimization intensity rises
rapidly with the increase of optimization dimensions, imposing higher demands on optimization methods. Effectively finding
the optimal solution in high-dimensional spaces has become a significant challenge. Therefore, this study innovatively
establishes an optimization framework combining deep learning and reinforcement learning, effectively solving high
dimensional optimization problems. Specifically, this framework proposes an intelligent optimization method that integrates
Deep Neural Network (DNN) prediction model and Proximal Policy Optimization (PPO) algorithm, and tests it on the
aerodynamic performance optimization problem of the airfoil NACA0012. It is compared with four traditional optimization
algorithms: genetic algorithm, particle swarm optimization, ant colony algorithm, and simulated annealing, under the setting
conditions of design parameter dimensions of 10, 50 and 100 respectively. The study shows that by comparing the optimization
effects of different algorithms, the optimization framework based on deep reinforcement learning outperforms traditional
optimization algorithms in airfoil examples from low to high dimensions, with optimization magnitudes of 16.02%, 50.08%, and
40% under the dimensions of 10, 50, and 100, respectively, and the impact of high dimensions is the smallest among the
algorithms compared. The results indicate that the established deep reinforcement learning framework has good application
prospects in high-dimensional optimization scenarios.
Key Words
airfoil Optimization; deep learning; high-dimensional optimization; reinforcement learning
Address
Ke Li: 1)Key Laboratory of New Technology for Construction of Cities in Mountain Area (Chongqing University),
Ministry of Education, Chongqing, China, 400045
2)School of Civil Engineering, Chongqing University, Chongqing, China, 400045
Haoyu Peng: School of Civil Engineering, Chongqing University, Chongqing, China, 400045
Zengshun Chen: 1)Key Laboratory of New Technology for Construction of Cities in Mountain Area (Chongqing University),
Ministry of Education, Chongqing, China, 400045
2)School of Civil Engineering, Chongqing University, Chongqing, China, 400045
Yi Hui: 1)Key Laboratory of New Technology for Construction of Cities in Mountain Area (Chongqing University),
Ministry of Education, Chongqing, China, 400045
2)School of Civil Engineering, Chongqing University, Chongqing, China, 400045
Abstract
This study investigates the aerodynamic influence of Swirl Recovery Vanes (SRVs) integrated with the SR3 rotor in
a transonic tractor propeller system. Computational simulations were performed using Analysis System Computational Fluid
Dynamics (ANSYS CFX) R to evaluate SRV configurations with ten and twelve blades, all designed with a National Advisory
Committee for Aeronautics (NACA) 4703 airfoil profile. The investigation was conducted at an altitude of 35,000 feet, with a
freestream Mach number of 0.8 and an advance ratio of 3.6. The findings indicate that increasing the SRV blade count enhances
swirl recovery, leading to improved axial velocity and reduced swirl losses. Among the examined configurations, the ten bladed
SRV setup demonstrated the highest overall efficiency, achieving a propeller thrust coefficient improvement of 5.09%. The
twelve bladed configuration also showed improved swirl recovery but did not surpass the efficiency gains observed with the ten
bladed setup. Additionally, increasing the SRV pitch angle contributed to an 8.76% enhancement in the propeller thrust
coefficient. These results underscore the significance of selecting an appropriate SRV blade count and pitch angle to maximize
propulsion efficiency in high-speed aircraft applications.
Key Words
power coefficient; propulsive efficiency; rotor; swirl recovery vane; thrust coefficient
Address
Santhi Raviselvam: Department of Aeronautical Engineering, Hindustan Institute of Technology and Science, Chennai 603103, India
Vasanthakumar Parthasarathy: Department of Aeronautical Engineering, Hindustan Institute of Technology and Science, Chennai 603103, India
Abstract
The precise and efficient simulation of nonstationary nonhomogeneous wind fields with intricate spatio-temporal
characteristics is the premise of wind-resistant analysis and design for large-scale wind turbine structures. However, when
generating nonstationary nonhomogeneous wind fields in two spatial dimensions with time-varying coherence function (TVCF)
based on the stochastic wave spectral representation method (SWSRM), significant computational challenges persist. The
construction, decomposition, and multifold summation operations of the high-dimensional evolutionary wavenumber-frequency
joint spectrum (EWFJS) matrix, characterized by the coupling of frequency, time, wavenumber, and height, demand excessive
memory allocation and exhibit inefficient computational performance. In this work, an innovative time-frequency interpolation
(TFI) enhanced SWSRM is proposed. This methodology commences by constructing the EWFJS matrix exclusively at the well
designed non-uniform time-frequency interpolation knots. Simultaneously, the hierarchical decomposition process, that is the
proper orthogonal decomposition-based multivariate decoupling method (POD-MDM), is proposed to systematically decouple
multiple variables in the high-dimensional EWFJS matrix. Subsequently, the tailored TFI scheme is developed to generate the
surrogate model which accurately represents the EWFJS matrix. This approach not only diminishes computational memory
usage through dimensionality reduction of EWFJS but also facilitates Fast Fourier Transform (FFT) implementation across time,
frequency and wavenumber dimensions, thereby leading to substantial improvements in computational performance. Numerical
examples in simulating nonstationary nonhomogeneous wind fields with TVCF of a wind turbine validate the accuracy and
computational efficiency, given the comparison of the spectrum and computational time between the traditional SWSRM and
the proposed TFI enhanced SWSRM.
Key Words
evolutionary wavenumber-frequency joint spectrum; multivariate hierarchical decomposition; nonstationary
nonhomogeneous wind field; time-frequency interpolation; time-varying coherence function
Address
Xinyi Huang: School of Mechanics and Engineering Science, Shanghai University, 200444 Shanghai, China
Feng Wang: Shanghai Institute of Applied Physics, Chinese Academy of Sciences, 201800 Shanghai, China
Chong Zhou: Shanghai Institute of Applied Physics, Chinese Academy of Sciences, 201800 Shanghai, China
Chunxiang Li: School of Mechanics and Engineering Science, Shanghai University, 200444 Shanghai, China
Liyuan Cao: School of Mechanics and Engineering Science, Shanghai University, 200444 Shanghai, China
