RESEARCH ON INTELLIGENT OPTIMIZATION OF FLOATING PLATFORM MAIN SCALE PARAMETERS FCR SEMI-SUBMERSIBLE WIND TURBINES
In this paper,four important main scale parameters are determined to express the platform design,and 120 sets of parameter combinations are formed.Based on OpenFAST and AQWA,the numerical simulation of the platform with all parameter combinations is carried out,and the short-term extreme value of motion response is predicted,forming a neural network training database.Secondly,the training of BP neural network model is completed,so that the short-term extreme prediction error of heave,pitch,yaw and yaw bearing fore-aft acceleration is less than 10%,and the surrogate model is formed.Finally,the genetic algorithm is used to optimize the heave motion response of the platform,and the extreme value of the heave response of the optimal platform scheme is further reduced by 6.98%compared with the lowest value in the database.This paper provides an intelligent method for the main scale parameter optimization of floating wind turbine platform.
NETL
NSTL
万方数据
