A reinforcement learning based blade twist angle distribution searching method for optimizing wind turbine energy power

Abstract The twist angle distribution (TAD) of a wind turbine blade determines its efficiency in terms of electricity production. As the blade is normally deployed in dynamic wind environments where wind speed varies widely, it is crucial to search the optimal TAD for different wind speeds in the design process. Due to the need to call a large number of complex simulators, traditional methods for optimal TAD searching are inefficient and time-consuming, which hinders the blade design process. Hence, this work presents a reinforcement learning-based method for searching optimal TAD efficiently, which is named RL-TAD. The fundamental idea of RL-TAD is to learn the TAD searching policy by an agent and reuse this agent to search optimal TAD under different wind speeds. This idea divided RL-TAD from the commonly used genetic algorithm-based methods, which only output the TAD and ignore reusing the searching policy. The RL-TAD includes the offline stage and online stage. In the offline stage, the RL-TAD learns the policy by reinforcement learning, in which the environment is constructed by a surrogate model, and a new reward policy is developed by integrating design experiences. Then in the online stage, the trained agent can be deployed to search TAD for different wind speeds. To verify the proposed RL-TAD, a case study is detailed. The empirical results show that the RL-TAD can converge to the optimal TAD at a speed of 3–5 times faster than the genetic algorithm-based method in the offline stage and also a better wind energy power coefficient is obtained. Besides, the response time is less than 0.1 s when using the trained agent to search the TAD, which proofs its potential to be used in rapid optimal TAD searching. Further, the rapid optimal TAD searching ability can support the real-time control of the wind turbine blade.

[1]  Mariusz Pawlak,et al.  Optimisation of wind turbine blades , 2005 .

[2]  Jiangtao Cheng Design of Wind Turbine Airfoils Based on Maximum Power Coefficient , 2010 .

[3]  Onur Mutlu,et al.  Self-Optimizing Memory Controllers: A Reinforcement Learning Approach , 2008, 2008 International Symposium on Computer Architecture.

[4]  Jun Yang,et al.  A multi-objective optimization for HAWT blades design by considering structural strength , 2016 .

[5]  Meng-Sing Liou,et al.  Transonic Axial-Flow Blade Optimization: Evolutionary Algorithms/Three-Dimensional Navier-Stokes Solver , 2004 .

[6]  Andrew Kusiak,et al.  Optimization of wind turbine energy and power factor with an evolutionary computation algorithm , 2010 .

[7]  Farrokh Mistree,et al.  A rule-based method for automated surrogate model selection , 2020, Adv. Eng. Informatics.

[8]  Yuan Luo,et al.  Improved non-dominated sorting genetic algorithm (NSGA)-II in multi-objective optimization studies of wind turbine blades , 2011 .

[9]  Hamid Khakpour Nejadkhaki,et al.  Modeling and Design Method for an Adaptive Wind Turbine Blade With Out-of-Plane Twist , 2018 .

[10]  M. Tahani,et al.  Aerodynamic design of horizontal axis wind turbine with innovative local linearization of chord and twist distributions , 2017 .

[11]  Lin Wang,et al.  Optimized linearization of chord and twist angle profiles for fixed-pitch fixed-speed wind turbine blades , 2013 .

[12]  Giacomo Frulla,et al.  A variable twist blade concept for more effective wind generation: design and realization , 2016 .

[13]  Demis Hassabis,et al.  Mastering the game of Go with deep neural networks and tree search , 2016, Nature.

[14]  Joaquim R. R. A. Martins,et al.  Aerostructural Shape Optimization of Wind Turbine Blades Considering Site-Specic Winds , 2008 .

[15]  Andy J. Keane,et al.  Robust Turbine Blade Optimization in the Face of Real Geometric Variations , 2018, Journal of Propulsion and Power.

[16]  Néstor Ramos-García,et al.  Design optimization of a curved wind turbine blade using neural networks and an aero-elastic vortex method under turbulent inflow , 2020 .

[17]  Ernesto Benini,et al.  Proposal for a coupled aerodynamic–structural wind turbine blade optimization , 2017 .

[18]  Kwang-Yong Kim,et al.  Multiple surrogate modeling for axial compressor blade shape optimization , 2008 .

[19]  Timoleon Kipouros,et al.  Multi-objective optimisation of horizontal axis wind turbine structure and energy production using aerofoil and blade properties as design variables , 2014 .

[20]  E. Lund,et al.  Structural gradient based sizing optimization of wind turbine blades with fixed outer geometry , 2018, Composite Structures.

[21]  Mahdi Moghimi Zand,et al.  Aerodynamic design and economical evaluation of site specific horizontal axis wind turbine (HAWT) , 2014 .

[22]  David Greiner,et al.  Wind blade chord and twist angle optimization using genetic algorithms , 2006 .

[23]  Tianyou Chai,et al.  Optimal Output Regulation of Linear Discrete-Time Systems With Unknown Dynamics Using Reinforcement Learning , 2020, IEEE Transactions on Cybernetics.

[24]  Wen Zhong Shen,et al.  Structural optimization study of composite wind turbine blade , 2013 .

[25]  W. Shen,et al.  Optimization design of blade shapes for wind turbines , 2010 .

[26]  Wei Qiao,et al.  An Adaptive Network-Based Reinforcement Learning Method for MPPT Control of PMSG Wind Energy Conversion Systems , 2016, IEEE Transactions on Power Electronics.

[27]  Alex Graves,et al.  Playing Atari with Deep Reinforcement Learning , 2013, ArXiv.

[28]  Alireza Maheri,et al.  Multiobjective optimisation and integrated design of wind turbine blades using WTBM-ANSYS for high fidelity structural analysis , 2020 .

[29]  Abhijit Gosavi,et al.  Reinforcement Learning: A Tutorial Survey and Recent Advances , 2009, INFORMS J. Comput..

[30]  Wang Xudong,et al.  Shape Optimization of Wind Turbine Blades , 2009 .

[31]  SangJoon Shin,et al.  Wind Turbine Blade Optimal Design Considering Multi-Parameters and Response Surface Method , 2020 .

[32]  Y. M. Wang,et al.  A reinforcement learning approach to dynamic optimization of load allocation in AGC system , 2009, 2009 IEEE Power & Energy Society General Meeting.

[33]  Dale Schuurmans,et al.  Bridging the Gap Between Value and Policy Based Reinforcement Learning , 2017, NIPS.

[34]  Jia Hao,et al.  Evolutionary Neural Network-based Method for Constructing Surrogate Model with Small Scattered Dataset and Monotonicity Experience , 2018, 2018 5th International Conference on Soft Computing & Machine Intelligence (ISCMI).

[35]  S. Report,et al.  The Sandia 100-meter All-glass Baseline Wind Turbine Blade: SNL100-00 , 2011 .

[36]  Michael S. Selig,et al.  Morphing Segmented Wind Turbine Concept , 2010 .

[37]  Kexiang Wei,et al.  Improving wind turbine blade based on multi-objective particle swarm optimization , 2020 .

[38]  Hamid Khakpour Nejadkhaki,et al.  Variable Twist Blade Transformation to Improve Wind Turbine Performance , 2018, Volume 6B: Energy.

[39]  Maureen Hand,et al.  Unsteady Aerodynamics Experiment Phase VI: Wind Tunnel Test Con gurations and Available Data Campaigns , 2001 .

[40]  Wang Yuming Mechanism of Combined Coning and Waviness Mechanical Face Seal for Nuclear Reactor Coolant Pump , 2010 .