Minimum Entropy Control?
For non-Gaussian stochastic systems, minimum entropy control methods are used to instead of the minimum variance control method in Gaussian systems. In this paper, some causes about the minimum entropy control may not be realized are given based on analysis the Shannon entropy definition of continuous random variable. On this basis, a new entropy, which has a similar nature with shannon entropy, construction method is presented, and some key properties of the new entropy are proved. Such nature of the new entropy shows that it meets the basic requirements of design of the minimum new entropy controller. Nextly, some conclusions are extended to the PDF by B-spline basis function approximation systems. And then, a minimum new entropy control with the mean constraint is constructed. Simulation illustrates the effectiveness of the minimum new entropy control with the mean constraint controller.
Stochastic control system Shannon Entropy Probability density function (PDF) Mean constrain B-spline basis function
ZHOU Jinglin ZHU Haijiang Wang Jing
College of Information Science and Technology, Beijing University of Chemical Technology, Beijing, China
国际会议
The 31st Chinese Control Conference(第三十一届中国控制会议)
合肥
英文
1590-1595
2012-07-01（万方平台首次上网日期，不代表论文的发表时间）