Study on PSO-BP-based Coupling Relationship of Axial Flow Rotor Turbine
The efficiency of Kaplan turbine units is closely related to the guide vane-paddle coupling relationship of the tur-bine.At present,the coupling relationship of hydraulic turbine is obtained by linear interpolation method,and its value is quite different from that of hydraulic turbine in optimal operating condition.In view of the shortcoming that the interpola-tion method cannot correctly reflect the optimal co-coupling relationship,this paper makes use of particle swarm neural network(PSO-BP)to train and test original relationship between guide vane and paddle,and obtains better coupling rela-tionship.Through the hydraulic turbine group test,the paddle opening value of PSO-BP algorithm is reduced by 3.61%on average than that of linear interpolation method,and the efficiency is improved by 3.09%.The experimental results show that the PSO-BP optimized guide vane-paddle coupling relationship can reduce overwater flow,save hydraulic resource and improve operating efficiency of the hydraulic turbine.