Study On Blade Shape Optimization Of S-Type Wind Turbine Based On Radial Basis Function Model And Marine Predator Algorithm

The use of traditional fossil energy sources has not only created a growing pollution problem for the environment,but has also had an irreversible negative impact on the ecosystem.Therefore,the use of renewable energy has become a major strategic initiative and universal consensus for countries around the world to seize the future energy heights,maintain national energy security and deal with ecological and environmental problems.With the growing wind power industry and the unique advantages of wind power,wind energy is considered to be one of the most promising renewable energy sources in the future.Savonius wind turbine is a machine that transforms wind energy into electricity,which has many advantages and promising applications,but the inherent defect of low energy capture efficiency limits its development and promotion.Blade as a critical part of wind turbine energy capture,optimize the blade shape can directly improve the energy capture efficiency of S-type wind turbine.In this thesis,we propose a set of efficient,intelligent and reliable S-shaped fan blade optimization design method for the problems of poor effect and low efficiency of traditional S-shaped wind turbine blade shape optimization methods.First,the shape of the blade is parametrically characterized using a third-order Bessel curve,then a certain number of design solutions are obtained in the design space using the Latin hypercube sampling method and the moment coefficients of each design solution are evaluated based on CFD,then an RBF proxy model is constructed based on these design solutions and their corresponding moment coefficients,and finally the optimal design solution is obtained by solving the proxy model using the marine predator algorithm and verified using CFD.This thesis further investigates a dynamic radial basis function proxy model constructed based on different initial sample schemes to further ensure the global approximation accuracy.The method is used to improve the generalization ability of the RBF model by continuously adding the optimal design solution found last time and the corresponding true value to the initial sample pool,and after several cycles,a better design solution can be found.This thesis systematically explains the reasons for the outstanding performance of the optimized blade compared to the conventional blade in terms of aerodynamic performance.For the disadvantage of the optimized blade in the return process with large negative moment,the wind collector and the improved wind collector are proposed as auxiliary devices to improve the disadvantage of the blade in the return process to further improve the performance of the S-type wind turbine.The results of the study show that the drag force on the blades during the return process can be effectively improved by adding an auxiliary device.The average torque coefficient of the S-type wind turbine blade obtained by the above optimization method can be increased by about 7% compared with the conventional S-type wind turbine blade at the blade tip speed ratio of 1.At the same time,the average torque coefficient of the S-type fan blade at other blade tip speed ratios has a significant increase compared with the conventional S-type wind turbine blade.At a blade tip speed ratio of 1,the moment coefficient of the blade with the addition of the wind collector is increased by about 41% compared to the optimized blade.The moment coefficient is improved by about 52% compared to the classical blade.The performance of the blades with the modified wind collector was improved by about 42% compared to the blades with the wind collector.This study verifies the applicability of radial basis function model and marine predator algorithm in the blade shape optimization problem of S-type wind turbine,and provides an intelligent optimization theory for the blade shape optimization problem of S-type wind turbine,and also provides some theoretical basis for the shape optimization problem of other complex mechanical components.The auxiliary devices studied in this thesis also provide data support for in-depth studies in subsequent work to reduce the workload and blindness of subsequent studies.