Recursive least-squares temporal difference algorithm (RLS-TD) is deduced, which can use data more efficiently with fast convergence and less computational burden. Reinforcement learning based on recursive least-squar...
详细信息
A method, based on Genetic Algorithm, for optimizing multivariable PID controllers, is being proposed. The global parameters of the PID controller are optimized here by utilizing the generic framework of genetic algor...
详细信息
A method, based on Genetic Algorithm, for optimizing multivariable PID controllers, is being proposed. The global parameters of the PID controller are optimized here by utilizing the generic framework of genetic algorithm. The only prerequisite asked for by the method is that the object function can be numerically evaluated, without any restrictions concerning the concrete form of the controlled object's model, and without requiring any analytical relationship between the object function and the control parameters as well. The method is therefore featured by superior universality. simulation of a steam turbine generator set' s controlsystem shows that, though the decoupling effect of multivariable PID controllers, designed by the mentioned method, is somewhat inferior to the optimization of parameterized high order normal matrix design method, but has, in respect to decoupling effect, remarkable improvements in comparison with multi-loop PLD control. It has, in addition, a certain disturbance resistant capability, and robustness;is exhibiting good dynamic behavior and proves the method's feasibility.
Recursive least-squares temporal difference algorithm (RLS-TD) is deduced, which can use data more efficiently with fast convergence and less computational burden. Reinforcement learning based on recursive least-squar...
详细信息
Recursive least-squares temporal difference algorithm (RLS-TD) is deduced, which can use data more efficiently with fast convergence and less computational burden. Reinforcement learning based on recursive least-squares methods is applied to ship steering control, as provides an efficient way for the improvement of ship steering control performance. It removes the defect that the conventional intelligent algorithm learning must be provided with some sample data. The parameters of controller are on-line learned and adjusted. simulation results show that the ship course can be properly controlled in case of the disturbances of wave, wind, current. It is demonstrated that the proposed algorithm is a promising alternative to conventional autopilots.
作者：
吴伟仁田玉龙黄翔宇Institute for Pattern Recognition and Artificial Intelligence
State Key Lab. for Image Processing and Intelligent Control Huazhong University of Science and Technology Wuhan 430074 China Deep Space Exploration Research Center
Harbin Institute of Technology Harbin 150001 Chinahe image elements of earth-center and moon-center are obtained by processing the images of earth and moon these image elements in combination with the inertial attitude information and the moon ephemeris are utilized to obtain the probe initial position relative to earth and the Levenberg-Marquardt algorithm is used to determine the accurate probe position relative to earth and the probe orbit relative to earth is estimated by using the extended Kalman filter. The autonomous optical navigation algorithm is validated using the digital simulation.
The image elements of earth-center and moon-center are obtained by processing the images of earthand moon, these image elements in combination with the inertial attitude information and the moon ephemerisare utilized ...
详细信息
The image elements of earth-center and moon-center are obtained by processing the images of earthand moon, these image elements in combination with the inertial attitude information and the moon ephemerisare utilized to obtain the probe initial position relative to earth, and the Levenberg-Marquardt algorithm is usedto determine the accurate probe position relative to earth, and the probe orbit relative to earth is estimated by u-sing the extended Kalman filter. The autonomous optical navigation algorithm is validated using the digital simu-lation.
暂无评论