A back propagation （BP） neural network mathematical model was established to investigate the maneuvering control of an air cushion vehicle （ACV）. The calculation was based on four-freedom-degree model experiments ...A back propagation （BP） neural network mathematical model was established to investigate the maneuvering control of an air cushion vehicle （ACV）. The calculation was based on four-freedom-degree model experiments of hydrodynamics and aerodynamics. It is necessary for the ACV to control the velocity and the yaw rate as well as the velocity angle at the same time. The yaw rate and the velocity angle must be controlled correspondingly because of the whipping, which is a special characteristic for the ACV. The calculation results show that it is an efficient way for the ACV＇s maneuvering control by using a BP neural network to adjust PID parameters online.展开更多
A new airfoil shape parameterization method is developed, which extended the Bezier curve to the generalized form with adjustable shape parameters. The local control parameters at airfoil leading and trailing edge reg...A new airfoil shape parameterization method is developed, which extended the Bezier curve to the generalized form with adjustable shape parameters. The local control parameters at airfoil leading and trailing edge regions are enhanced, where have significant effect on the aerodynamic performance of wind turbine. The results show this improved parameterization method has advantages in the fitting characteristics of geometry shape and aero- dynamic performance comparing with other three common airfoil parameterization methods. The new paramete- rization method is then applied to airfoil shape optimization for wind turbine using Genetic Algorithm （GA）, and the wind turbine special airfoil, DU93-W-210, is optimized to achieve the favorable C1/Cd at specified flow con- ditions. The aerodynamic characteristic of the optimum airfoil is obtained by solving the RANS equations in computational fluid dynamics （CFD） method, and the optimization convergence curves show that the new para- meterization method has good convergence rate in less number of generations comparing with other methods. It is concluded that the new method not only has well controllability and completeness in airfoil shape representation and provides more flexibility in expressing the airfoil geometry shape, but also is capable to find efficient and op- timal wind turbine airfoil. Additionally, it is shown that a suitable parameterization method is helpful for improv- ing the convergence rate of the optimization algorithm.展开更多
文摘A back propagation （BP） neural network mathematical model was established to investigate the maneuvering control of an air cushion vehicle （ACV）. The calculation was based on four-freedom-degree model experiments of hydrodynamics and aerodynamics. It is necessary for the ACV to control the velocity and the yaw rate as well as the velocity angle at the same time. The yaw rate and the velocity angle must be controlled correspondingly because of the whipping, which is a special characteristic for the ACV. The calculation results show that it is an efficient way for the ACV＇s maneuvering control by using a BP neural network to adjust PID parameters online.
基金funded by the National Natural Science Foundation of China(No.51376024)the Specialized Research Fund for the Doctoral Program of Higher Education(No.20131101110015),China
文摘A new airfoil shape parameterization method is developed, which extended the Bezier curve to the generalized form with adjustable shape parameters. The local control parameters at airfoil leading and trailing edge regions are enhanced, where have significant effect on the aerodynamic performance of wind turbine. The results show this improved parameterization method has advantages in the fitting characteristics of geometry shape and aero- dynamic performance comparing with other three common airfoil parameterization methods. The new paramete- rization method is then applied to airfoil shape optimization for wind turbine using Genetic Algorithm （GA）, and the wind turbine special airfoil, DU93-W-210, is optimized to achieve the favorable C1/Cd at specified flow con- ditions. The aerodynamic characteristic of the optimum airfoil is obtained by solving the RANS equations in computational fluid dynamics （CFD） method, and the optimization convergence curves show that the new para- meterization method has good convergence rate in less number of generations comparing with other methods. It is concluded that the new method not only has well controllability and completeness in airfoil shape representation and provides more flexibility in expressing the airfoil geometry shape, but also is capable to find efficient and op- timal wind turbine airfoil. Additionally, it is shown that a suitable parameterization method is helpful for improv- ing the convergence rate of the optimization algorithm.
