期刊文献+

重力训练法

Gravity Training Method
下载PDF
导出
摘要 针对某些对象梯度不易计算的问题,提出一种新的优化训练方法.利用水往低处流的原理对参数进行寻优,将随机初始参数根据步长生成多维空间中点的坐标形式,用实际输出和目标输出之间的差来表示以此点为球心的球的半径,将最小半径的球作为下一步寻优的中心点.此方法具有无需计算梯度,初始值随机选取,易编程,寻优快等特点.通过仿真实验,此方法成功应用在PID控制器、LQR控制器和神经网络中,获取了最优参数,具有实际操作性. A newtraining method was proposed to solve the problem that some gradients of the objects are not easy to calculate. This method was based on the principle of gravity optimizes parameters. Random initial parameter based on step was set as coordinate form which in the midpoint of the multidimensional space. The error between the actual output and the target output was set as radius. This method had advantages which could not need to calculate the gradient and could randomly select initial value. This method was successfully used in the PID controller,LQR controller and neural network.
出处 《东北大学学报(自然科学版)》 EI CAS CSCD 北大核心 2015年第11期1558-1561,共4页 Journal of Northeastern University(Natural Science)
基金 辽宁省自然科学基金资助项目(20102127) 辽宁省高校创新团队支持计划项目(2009T062.LT2010058)
关键词 重力 训练 优化 梯度 神经网络 PID控制器 LQR控制器 参数 gravity training optimization gradient neural network PID controller LQR controller parameter
  • 引文网络
  • 相关文献

参考文献7

  • 1Xu Y R, Ford J, Becker E, et al. A BP-neural network improvement to hop-counting for localization in wireless sensor networks I M ~//Tools and Applications with Artificial Intelligence. Berlin: Springer Berlin Heidelberg, 2009 : 11 - 23.
  • 2徐林,张宇献,王建辉,顾树生.基于多值编码GA-BP混合算法的板形板厚综合预测控制[J].东南大学学报(自然科学版),2005,35(A02):132-136. 被引量:1
  • 3Pendharkar P C. A comparison of gradient ascent, gradient descent and genetic-algorithm-based artificial neural networks for the binary classification problem ~ J ]. Expert Systems, 2007.24(2) :65 -86.
  • 4侯嫒彬,杜京义,汪梅.神经网络[M].西安:西安电子科技大学出版社,2007:15—25.
  • 5董海鹰.智能控制理论及应用[M].北京:中国铁道出版社.2006.70—85.
  • 6Zhang B L, Wang J G. The analysis and simulation of first- order inverted pendulum control system based on Lqr [ C ]// 3rd International Symposium on Information Processing. Washington D C ,2010:447 - 449.
  • 7Yu J, Fang J. Inverted pendulum RBF neural network PID controller design [ C ]// IEEE 2014 International Symposium on Computer, Consumer and Control (IS3C). Taichung, 2014:560 - 562.

二级参考文献7

共引文献12

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部